Unnamed: 0
int64 0
5k
| title
stringlengths 9
210
| abstract
stringlengths 164
1.92k
|
|---|---|---|
2,500 | Scalable Online Disease Diagnosis via Multi-Model-Fused Actor-Critic Reinforcement Learning | For those seeking healthcare advice online, AI based dialogue agents capable
of interacting with patients to perform automatic disease diagnosis are a
viable option. This application necessitates efficient inquiry of relevant
disease symptoms in order to make accurate diagnosis recommendations. This can
be formulated as a problem of sequential feature (symptom) selection and
classification for which reinforcement learning (RL) approaches have been
proposed as a natural solution. They perform well when the feature space is
small, that is, the number of symptoms and diagnosable disease categories is
limited, but they frequently fail in assignments with a large number of
features. To address this challenge, we propose a Multi-Model-Fused
Actor-Critic (MMF-AC) RL framework that consists of a generative actor network
and a diagnostic critic network. The actor incorporates a Variational
AutoEncoder (VAE) to model the uncertainty induced by partial observations of
features, thereby facilitating in making appropriate inquiries. In the critic
network, a supervised diagnosis model for disease predictions is involved to
precisely estimate the state-value function. Furthermore, inspired by the
medical concept of differential diagnosis, we combine the generative and
diagnosis models to create a novel reward shaping mechanism to address the
sparse reward problem in large search spaces. We conduct extensive experiments
on both synthetic and real-world datasets for empirical evaluations. The
results demonstrate that our approach outperforms state-of-the-art methods in
terms of diagnostic accuracy and interaction efficiency while also being more
effectively scalable to large search spaces. Besides, our method is adaptable
to both categorical and continuous features, making it ideal for online
applications. |
2,501 | Joint Adversarial Learning for Cross-domain Fair Classification | Modern machine learning (ML) models are becoming increasingly popular and are
widely used in decision-making systems. However, studies have shown critical
issues of ML discrimination and unfairness, which hinder their adoption on
high-stake applications. Recent research on fair classifiers has drawn
significant attention to develop effective algorithms to achieve fairness and
good classification performance. Despite the great success of these
fairness-aware machine learning models, most of the existing models require
sensitive attributes to preprocess the data, regularize the model learning or
postprocess the prediction to have fair predictions. However, sensitive
attributes are often incomplete or even unavailable due to privacy, legal or
regulation restrictions. Though we lack the sensitive attribute for training a
fair model in the target domain, there might exist a similar domain that has
sensitive attributes. Thus, it is important to exploit auxiliary information
from the similar domain to help improve fair classification in the target
domain. Therefore, in this paper, we study a novel problem of exploring domain
adaptation for fair classification. We propose a new framework that can
simultaneously estimate the sensitive attributes while learning a fair
classifier in the target domain. Extensive experiments on real-world datasets
illustrate the effectiveness of the proposed model for fair classification,
even when no sensitive attributes are available in the target domain. |
2,502 | pFL-Bench: A Comprehensive Benchmark for Personalized Federated Learning | Personalized Federated Learning (pFL), which utilizes and deploys distinct
local models, has gained increasing attention in recent years due to its
success in handling the statistical heterogeneity of FL clients. However,
standardized evaluation and systematical analysis of diverse pFL methods remain
a challenge. Firstly, the highly varied datasets, FL simulation settings and
pFL implementations prevent fast and fair comparisons of pFL methods. Secondly,
the effectiveness and robustness of pFL methods are under-explored in various
practical scenarios, such as new clients generalization and resource-limited
clients participation. Finally, the current pFL literature diverges in the
adopted evaluation and ablation protocols. To tackle these challenges, we
propose the first comprehensive pFL benchmark, pFL-Bench, for facilitating
rapid, reproducible, standardized and thorough pFL evaluation. The proposed
benchmark contains more than 10 datasets in diverse application domains with
unified data partition and realistic heterogeneous settings; a modular and
easy-to-extend pFL codebase with more than 20 competitive pFL baseline
implementations; and systematic evaluations under containerized environments in
terms of generalization, fairness, system overhead, and convergence. We
highlight the benefits and potential of state-of-the-art pFL methods and hope
pFL-Bench enables further pFL research and broad applications that would
otherwise be difficult owing to the absence of a dedicated benchmark. The code
is released at
https://github.com/alibaba/FederatedScope/tree/master/benchmark/pFL-Bench. |
2,503 | Solving the Spike Feature Information Vanishing Problem in Spiking Deep Q Network with Potential Based Normalization | Brain inspired spiking neural networks (SNNs) have been successfully applied
to many pattern recognition domains. The SNNs based deep structure have
achieved considerable results in perceptual tasks, such as image
classification, target detection. However, the application of deep SNNs in
reinforcement learning (RL) tasks is still a problem to be explored. Although
there have been previous studies on the combination of SNNs and RL, most of
them focus on robotic control problems with shallow networks or using ANN-SNN
conversion method to implement spiking deep Q Network (SDQN). In this work, we
mathematically analyzed the problem of the disappearance of spiking signal
features in SDQN and proposed a potential based layer normalization(pbLN)
method to directly train spiking deep Q networks. Experiment shows that
compared with state-of-art ANN-SNN conversion method and other SDQN works, the
proposed pbLN spiking deep Q networks (PL-SDQN) achieved better performance on
Atari game tasks. |
2,504 | Neural Bandit with Arm Group Graph | Contextual bandits aim to identify among a set of arms the optimal one with
the highest reward based on their contextual information. Motivated by the fact
that the arms usually exhibit group behaviors and the mutual impacts exist
among groups, we introduce a new model, Arm Group Graph (AGG), where the nodes
represent the groups of arms and the weighted edges formulate the correlations
among groups. To leverage the rich information in AGG, we propose a bandit
algorithm, AGG-UCB, where the neural networks are designed to estimate rewards,
and we propose to utilize graph neural networks (GNN) to learn the
representations of arm groups with correlations. To solve the
exploitation-exploration dilemma in bandits, we derive a new upper confidence
bound (UCB) built on neural networks (exploitation) for exploration.
Furthermore, we prove that AGG-UCB can achieve a near-optimal regret bound with
over-parameterized neural networks, and provide the convergence analysis of GNN
with fully-connected layers which may be of independent interest. In the end,
we conduct extensive experiments against state-of-the-art baselines on multiple
public data sets, showing the effectiveness of the proposed algorithm. |
2,505 | Alternately Optimized Graph Neural Networks | Graph Neural Networks (GNNs) have demonstrated powerful representation
capability in numerous graph-based tasks. Specifically, the decoupled
structures of GNNs such as APPNP become popular due to their simplicity and
performance advantages. However, the end-to-end training of these GNNs makes
them inefficient in computation and memory consumption. In order to deal with
these limitations, in this work, we propose an alternating optimization
framework for graph neural networks that does not require end-to-end training.
Extensive experiments under different settings demonstrate that the performance
of the proposed algorithm is comparable to existing state-of-the-art algorithms
but has significantly better computation and memory efficiency. Additionally,
we show that our framework can be taken advantage to enhance existing decoupled
GNNs. |
2,506 | Network Report: A Structured Description for Network Datasets | The rapid development of network science and technologies depends on
shareable datasets. Currently, there is no standard practice for reporting and
sharing network datasets. Some network dataset providers only share links,
while others provide some contexts or basic statistics. As a result, critical
information may be unintentionally dropped, and network dataset consumers may
misunderstand or overlook critical aspects. Inappropriately using a network
dataset can lead to severe consequences (e.g., discrimination) especially when
machine learning models on networks are deployed in high-stake domains.
Challenges arise as networks are often used across different domains (e.g.,
network science, physics, etc) and have complex structures. To facilitate the
communication between network dataset providers and consumers, we propose
network report. A network report is a structured description that summarizes
and contextualizes a network dataset. Network report extends the idea of
dataset reports (e.g., Datasheets for Datasets) from prior work with
network-specific descriptions of the non-i.i.d. nature, demographic
information, network characteristics, etc. We hope network reports encourage
transparency and accountability in network research and development across
different fields. |
2,507 | Ensembles for Uncertainty Estimation: Benefits of Prior Functions and Bootstrapping | In machine learning, an agent needs to estimate uncertainty to efficiently
explore and adapt and to make effective decisions. A common approach to
uncertainty estimation maintains an ensemble of models. In recent years,
several approaches have been proposed for training ensembles, and conflicting
views prevail with regards to the importance of various ingredients of these
approaches. In this paper, we aim to address the benefits of two ingredients --
prior functions and bootstrapping -- which have come into question. We show
that prior functions can significantly improve an ensemble agent's joint
predictions across inputs and that bootstrapping affords additional benefits if
the signal-to-noise ratio varies across inputs. Our claims are justified by
both theoretical and experimental results. |
2,508 | Can Backdoor Attacks Survive Time-Varying Models? | Backdoors are powerful attacks against deep neural networks (DNNs). By
poisoning training data, attackers can inject hidden rules (backdoors) into
DNNs, which only activate on inputs containing attack-specific triggers. While
existing work has studied backdoor attacks on a variety of DNN models, they
only consider static models, which remain unchanged after initial deployment.
In this paper, we study the impact of backdoor attacks on a more realistic
scenario of time-varying DNN models, where model weights are updated
periodically to handle drifts in data distribution over time. Specifically, we
empirically quantify the "survivability" of a backdoor against model updates,
and examine how attack parameters, data drift behaviors, and model update
strategies affect backdoor survivability. Our results show that one-shot
backdoor attacks (i.e., only poisoning training data once) do not survive past
a few model updates, even when attackers aggressively increase trigger size and
poison ratio. To stay unaffected by model update, attackers must continuously
introduce corrupted data into the training pipeline. Together, these results
indicate that when models are updated to learn new data, they also "forget"
backdoors as hidden, malicious features. The larger the distribution shift
between old and new training data, the faster backdoors are forgotten.
Leveraging these insights, we apply a smart learning rate scheduler to further
accelerate backdoor forgetting during model updates, which prevents one-shot
backdoors from surviving past a single model update. |
2,509 | Subject Granular Differential Privacy in Federated Learning | This paper introduces subject granular privacy in the Federated Learning (FL)
setting, where a subject is an individual whose private information is embodied
by several data items either confined within a single federation user or
distributed across multiple federation users. We formally define the notion of
subject level differential privacy for FL. We propose three new algorithms that
enforce subject level DP. Two of these algorithms are based on notions of user
level local differential privacy (LDP) and group differential privacy
respectively. The third algorithm is based on a novel idea of hierarchical
gradient averaging (HiGradAvgDP) for subjects participating in a training
mini-batch. We also introduce horizontal composition of privacy loss for a
subject across multiple federation users. We show that horizontal composition
is equivalent to sequential composition in the worst case. We prove the subject
level DP guarantee for all our algorithms and empirically analyze them using
the FEMNIST and Shakespeare datasets. Our evaluation shows that, of our three
algorithms, HiGradAvgDP delivers the best model performance, approaching that
of a model trained using a DP-SGD based algorithm that provides a weaker item
level privacy guarantee. |
2,510 | Predictive Modeling of Charge Levels for Battery Electric Vehicles using CNN EfficientNet and IGTD Algorithm | Convolutional Neural Networks (CNN) have been a good solution for
understanding a vast image dataset. As the increased number of battery-equipped
electric vehicles is flourishing globally, there has been much research on
understanding which charge levels electric vehicle drivers would choose to
charge their vehicles to get to their destination without any prevention. We
implemented deep learning approaches to analyze the tabular datasets to
understand their state of charge and which charge levels they would choose. In
addition, we implemented the Image Generator for Tabular Dataset algorithm to
utilize tabular datasets as image datasets to train convolutional neural
networks. Also, we integrated other CNN architecture such as EfficientNet to
prove that CNN is a great learner for reading information from images that were
converted from the tabular dataset, and able to predict charge levels for
battery-equipped electric vehicles. We also evaluated several optimization
methods to enhance the learning rate of the models and examined further
analysis on improving the model architecture. |
2,511 | FedPop: A Bayesian Approach for Personalised Federated Learning | Personalised federated learning (FL) aims at collaboratively learning a
machine learning model taylored for each client. Albeit promising advances have
been made in this direction, most of existing approaches works do not allow for
uncertainty quantification which is crucial in many applications. In addition,
personalisation in the cross-device setting still involves important issues,
especially for new clients or those having small number of observations. This
paper aims at filling these gaps. To this end, we propose a novel methodology
coined FedPop by recasting personalised FL into the population modeling
paradigm where clients' models involve fixed common population parameters and
random effects, aiming at explaining data heterogeneity. To derive convergence
guarantees for our scheme, we introduce a new class of federated stochastic
optimisation algorithms which relies on Markov chain Monte Carlo methods.
Compared to existing personalised FL methods, the proposed methodology has
important benefits: it is robust to client drift, practical for inference on
new clients, and above all, enables uncertainty quantification under mild
computational and memory overheads. We provide non-asymptotic convergence
guarantees for the proposed algorithms and illustrate their performances on
various personalised federated learning tasks. |
2,512 | Towards Scalable Hyperbolic Neural Networks using Taylor Series Approximations | Hyperbolic networks have shown prominent improvements over their Euclidean
counterparts in several areas involving hierarchical datasets in various
domains such as computer vision, graph analysis, and natural language
processing. However, their adoption in practice remains restricted due to (i)
non-scalability on accelerated deep learning hardware, (ii) vanishing gradients
due to the closure of hyperbolic space, and (iii) information loss due to
frequent mapping between local tangent space and fully hyperbolic space. To
tackle these issues, we propose the approximation of hyperbolic operators using
Taylor series expansions, which allows us to reformulate the computationally
expensive tangent and cosine hyperbolic functions into their polynomial
equivariants which are more efficient. This allows us to retain the benefits of
preserving the hierarchical anatomy of the hyperbolic space, while maintaining
the scalability over current accelerated deep learning infrastructure. The
polynomial formulation also enables us to utilize the advancements in Euclidean
networks such as gradient clipping and ReLU activation to avoid vanishing
gradients and remove errors due to frequent switching between tangent space and
hyperbolic space. Our empirical evaluation on standard benchmarks in the domain
of graph analysis and computer vision shows that our polynomial formulation is
as scalable as Euclidean architectures, both in terms of memory and time
complexity, while providing results as effective as hyperbolic models.
Moreover, our formulation also shows a considerable improvement over its
baselines due to our solution to vanishing gradients and information loss. |
2,513 | Decoupled Self-supervised Learning for Non-Homophilous Graphs | In this paper, we study the problem of conducting self-supervised learning
for node representation learning on non-homophilous graphs. Existing
self-supervised learning methods typically assume the graph is homophilous
where linked nodes often belong to the same class or have similar features.
However, such assumptions of homophily do not always hold true in real-world
graphs. We address this problem by developing a decoupled self-supervised
learning (DSSL) framework for graph neural networks. DSSL imitates a generative
process of nodes and links from latent variable modeling of the semantic
structure, which decouples different underlying semantics between different
neighborhoods into the self-supervised node learning process. Our DSSL
framework is agnostic to the encoders and does not need prefabricated
augmentations, thus is flexible to different graphs. To effectively optimize
the framework with latent variables, we derive the evidence lower-bound of the
self-supervised objective and develop a scalable training algorithm with
variational inference. We provide a theoretical analysis to justify that DSSL
enjoys better downstream performance. Extensive experiments on various types of
graph benchmarks demonstrate that our proposed framework can significantly
achieve better performance compared with competitive self-supervised learning
baselines. |
2,514 | One Ring to Bring Them All: Towards Open-Set Recognition under Domain Shift | In this paper, we investigate $\textit{open-set recognition}$ with domain
shift, where the final goal is to achieve $\textit{Source-free Universal Domain
Adaptation}$ (SF-UNDA), which addresses the situation where there exist both
domain and category shifts between source and target domains. Under the SF-UNDA
setting, the model cannot access source data anymore during target adaptation,
which aims to address data privacy concerns. We propose a novel training scheme
to learn a ($n$+1)-way classifier to predict the $n$ source classes and the
unknown class, where samples of only known source categories are available for
training. Furthermore, for target adaptation, we simply adopt a weighted
entropy minimization to adapt the source pretrained model to the unlabeled
target domain without source data. In experiments, we show: $\textbf{1)}$ After
source training, the resulting source model can get excellent performance for
$\textit{open-set single domain generalization}$ and also $\textit{open-set
recognition}$ tasks; $\textbf{2)}$ After target adaptation, our method
surpasses current UNDA approaches which demand source data during adaptation on
several benchmarks. The versatility to several different tasks strongly proves
the efficacy and generalization ability of our method. $\textbf{3)}$ When
augmented with a closed-set domain adaptation approach during target
adaptation, our source-free method further outperforms the current
state-of-the-art UNDA method by 2.5%, 7.2% and 13% on Office-31, Office-Home
and VisDA respectively. Code will be available in
https://github.com/Albert0147/OneRing. |
2,515 | Meta-Learning Transferable Parameterized Skills | We propose a novel parameterized skill-learning algorithm that aims to learn
transferable parameterized skills and synthesize them into a new action space
that supports efficient learning in long-horizon tasks. We first propose novel
learning objectives -- trajectory-centric diversity and smoothness -- that
allow an agent to meta-learn reusable parameterized skills. Our agent can use
these learned skills to construct a temporally-extended parameterized-action
Markov decision process, for which we propose a hierarchical actor-critic
algorithm that aims to efficiently learn a high-level control policy with the
learned skills. We empirically demonstrate that the proposed algorithms enable
an agent to solve a complicated long-horizon obstacle-course environment. |
2,516 | Neural Network Compression via Effective Filter Analysis and Hierarchical Pruning | Network compression is crucial to making the deep networks to be more
efficient, faster, and generalizable to low-end hardware. Current network
compression methods have two open problems: first, there lacks a theoretical
framework to estimate the maximum compression rate; second, some layers may get
over-prunned, resulting in significant network performance drop. To solve these
two problems, this study propose a gradient-matrix singularity analysis-based
method to estimate the maximum network redundancy. Guided by that maximum rate,
a novel and efficient hierarchical network pruning algorithm is developed to
maximally condense the neuronal network structure without sacrificing network
performance. Substantial experiments are performed to demonstrate the efficacy
of the new method for pruning several advanced convolutional neural network
(CNN) architectures. Compared to existing pruning methods, the proposed pruning
algorithm achieved state-of-the-art performance. At the same or similar
compression ratio, the new method provided the highest network prediction
accuracy as compared to other methods. |
2,517 | Click prediction boosting via Bayesian hyperparameter optimization based ensemble learning pipelines | Online travel agencies (OTA's) advertise their website offers on meta-search
bidding engines. The problem of predicting the number of clicks a hotel would
receive for a given bid amount is an important step in the management of an
OTA's advertisement campaign on a meta-search engine, because bid times number
of clicks defines the cost to be generated. Various regressors are ensembled in
this work to improve click prediction performance. Following the preprocessing
procedures, the feature set is divided into train and test groups depending on
the logging date of the samples. The data collection is then subjected to
feature elimination via utilizing XGBoost, which significantly reduces the
dimension of features. The optimum hyper-parameters are then found by applying
Bayesian hyperparameter optimization to XGBoost, LightGBM, and SGD models. The
different trained models are tested separately as well as combined to form
ensemble models. Four alternative ensemble solutions have been suggested. The
same test set is used to test both individual and ensemble models, and the
results of 46 model combinations demonstrate that stack ensemble models yield
the desired R2 score of all. In conclusion, the ensemble model improves the
prediction performance by about 10%. |
2,518 | Spatial Cross-Attention Improves Self-Supervised Visual Representation Learning | Unsupervised representation learning methods like SwAV are proved to be
effective in learning visual semantics of a target dataset. The main idea
behind these methods is that different views of a same image represent the same
semantics. In this paper, we further introduce an add-on module to facilitate
the injection of the knowledge accounting for spatial cross correlations among
the samples. This in turn results in distilling intra-class information
including feature level locations and cross similarities between same-class
instances. The proposed add-on can be added to existing methods such as the
SwAV. We can later remove the add-on module for inference without any
modification of the learned weights. Through an extensive set of empirical
evaluations, we verify that our method yields an improved performance in
detecting the class activation maps, top-1 classification accuracy, and
down-stream tasks such as object detection, with different configuration
settings. |
2,519 | Distributed Newton-Type Methods with Communication Compression and Bernoulli Aggregation | Despite their high computation and communication costs, Newton-type methods
remain an appealing option for distributed training due to their robustness
against ill-conditioned convex problems. In this work, we study ommunication
compression and aggregation mechanisms for curvature information in order to
reduce these costs while preserving theoretically superior local convergence
guarantees. We prove that the recently developed class of three point
compressors (3PC) of Richtarik et al. [2022] for gradient communication can be
generalized to Hessian communication as well. This result opens up a wide
variety of communication strategies, such as contractive compression} and lazy
aggregation, available to our disposal to compress prohibitively costly
curvature information. Moreover, we discovered several new 3PC mechanisms, such
as adaptive thresholding and Bernoulli aggregation, which require reduced
communication and occasional Hessian computations. Furthermore, we extend and
analyze our approach to bidirectional communication compression and partial
device participation setups to cater to the practical considerations of
applications in federated learning. For all our methods, we derive fast
condition-number-independent local linear and/or superlinear convergence rates.
Finally, with extensive numerical evaluations on convex optimization problems,
we illustrate that our designed schemes achieve state-of-the-art communication
complexity compared to several key baselines using second-order information. |
2,520 | Certifying Data-Bias Robustness in Linear Regression | Datasets typically contain inaccuracies due to human error and societal
biases, and these inaccuracies can affect the outcomes of models trained on
such datasets. We present a technique for certifying whether linear regression
models are pointwise-robust to label bias in the training dataset, i.e.,
whether bounded perturbations to the labels of a training dataset result in
models that change the prediction of test points. We show how to solve this
problem exactly for individual test points, and provide an approximate but more
scalable method that does not require advance knowledge of the test point. We
extensively evaluate both techniques and find that linear models -- both
regression- and classification-based -- often display high levels of
bias-robustness. However, we also unearth gaps in bias-robustness, such as high
levels of non-robustness for certain bias assumptions on some datasets.
Overall, our approach can serve as a guide for when to trust, or question, a
model's output. |
2,521 | Overcoming the Long Horizon Barrier for Sample-Efficient Reinforcement Learning with Latent Low-Rank Structure | The practicality of reinforcement learning algorithms has been limited due to
poor scaling with respect to the problem size, as the sample complexity of
learning an $\epsilon$-optimal policy is $\Tilde{\Omega}\left(|S||A|H^3 /
\eps^2\right)$ over worst case instances of an MDP with state space $S$, action
space $A$, and horizon $H$. We consider a class of MDPs that exhibit low rank
structure, where the latent features are unknown. We argue that a natural
combination of value iteration and low-rank matrix estimation results in an
estimation error that grows doubly exponentially in the horizon $H$. We then
provide a new algorithm along with statistical guarantees that efficiently
exploits low rank structure given access to a generative model, achieving a
sample complexity of
$\Tilde{O}\left(d^5(|S|+|A|)\mathrm{poly}(H)/\eps^2\right)$ for a rank $d$
setting, which is minimax optimal with respect to the scaling of $|S|, |A|$,
and $\eps$. In contrast to literature on linear and low-rank MDPs, we do not
require a known feature mapping, our algorithm is computationally simple, and
our results hold for long time horizons. Our results provide insights on the
minimal low-rank structural assumptions required on the MDP with respect to the
transition kernel versus the optimal action-value function. |
2,522 | A generative recommender system with GMM prior for cancer drug generation and sensitivity prediction | Recent emergence of high-throughput drug screening assays sparkled an
intensive development of machine learning methods, including models for
prediction of sensitivity of cancer cell lines to anti-cancer drugs, as well as
methods for generation of potential drug candidates. However, a concept of
generation of compounds with specific properties and simultaneous modeling of
their efficacy against cancer cell lines has not been comprehensively explored.
To address this need, we present VADEERS, a Variational Autoencoder-based Drug
Efficacy Estimation Recommender System. The generation of compounds is
performed by a novel variational autoencoder with a semi-supervised Gaussian
Mixture Model (GMM) prior. The prior defines a clustering in the latent space,
where the clusters are associated with specific drug properties. In addition,
VADEERS is equipped with a cell line autoencoder and a sensitivity prediction
network. The model combines data for SMILES string representations of
anti-cancer drugs, their inhibition profiles against a panel of protein
kinases, cell lines biological features and measurements of the sensitivity of
the cell lines to the drugs. The evaluated variants of VADEERS achieve a high
r=0.87 Pearson correlation between true and predicted drug sensitivity
estimates. We train the GMM prior in such a way that the clusters in the latent
space correspond to a pre-computed clustering of the drugs by their inhibitory
profiles. We show that the learned latent representations and new generated
data points accurately reflect the given clustering. In summary, VADEERS offers
a comprehensive model of drugs and cell lines properties and relationships
between them, as well as a guided generation of novel compounds. |
2,523 | Extending Momentum Contrast with Cross Similarity Consistency Regularization | Contrastive self-supervised representation learning methods maximize the
similarity between the positive pairs, and at the same time tend to minimize
the similarity between the negative pairs. However, in general the interplay
between the negative pairs is ignored as they do not put in place special
mechanisms to treat negative pairs differently according to their specific
differences and similarities. In this paper, we present Extended Momentum
Contrast (XMoCo), a self-supervised representation learning method founded upon
the legacy of the momentum-encoder unit proposed in the MoCo family
configurations. To this end, we introduce a cross consistency regularization
loss, with which we extend the transformation consistency to dissimilar images
(negative pairs). Under the cross consistency regularization rule, we argue
that semantic representations associated with any pair of images (positive or
negative) should preserve their cross-similarity under pretext transformations.
Moreover, we further regularize the training loss by enforcing a uniform
distribution of similarity over the negative pairs across a batch. The proposed
regularization can easily be added to existing self-supervised learning
algorithms in a plug-and-play fashion. Empirically, we report a competitive
performance on the standard Imagenet-1K linear head classification benchmark.
In addition, by transferring the learned representations to common downstream
tasks, we show that using XMoCo with the prevalently utilized augmentations can
lead to improvements in the performance of such tasks. We hope the findings of
this paper serve as a motivation for researchers to take into consideration the
important interplay among the negative examples in self-supervised learning. |
2,524 | NOMAD: Nonlinear Manifold Decoders for Operator Learning | Supervised learning in function spaces is an emerging area of machine
learning research with applications to the prediction of complex physical
systems such as fluid flows, solid mechanics, and climate modeling. By directly
learning maps (operators) between infinite dimensional function spaces, these
models are able to learn discretization invariant representations of target
functions. A common approach is to represent such target functions as linear
combinations of basis elements learned from data. However, there are simple
scenarios where, even though the target functions form a low dimensional
submanifold, a very large number of basis elements is needed for an accurate
linear representation. Here we present NOMAD, a novel operator learning
framework with a nonlinear decoder map capable of learning finite dimensional
representations of nonlinear submanifolds in function spaces. We show this
method is able to accurately learn low dimensional representations of solution
manifolds to partial differential equations while outperforming linear models
of larger size. Additionally, we compare to state-of-the-art operator learning
methods on a complex fluid dynamics benchmark and achieve competitive
performance with a significantly smaller model size and training cost. |
2,525 | A Penny for Your (visual) Thoughts: Self-Supervised Reconstruction of Natural Movies from Brain Activity | Reconstructing natural videos from fMRI brain recordings is very challenging,
for two main reasons: (i) As fMRI data acquisition is difficult, we only have a
limited amount of supervised samples, which is not enough to cover the huge
space of natural videos; and (ii) The temporal resolution of fMRI recordings is
much lower than the frame rate of natural videos. In this paper, we propose a
self-supervised approach for natural-movie reconstruction. By employing
cycle-consistency over Encoding-Decoding natural videos, we can: (i) exploit
the full framerate of the training videos, and not be limited only to clips
that correspond to fMRI recordings; (ii) exploit massive amounts of external
natural videos which the subjects never saw inside the fMRI machine. These
enable increasing the applicable training data by several orders of magnitude,
introducing natural video priors to the decoding network, as well as temporal
coherence. Our approach significantly outperforms competing methods, since
those train only on the limited supervised data. We further introduce a new and
simple temporal prior of natural videos, which - when folded into our fMRI
decoder further - allows us to reconstruct videos at a higher frame-rate (HFR)
of up to x8 of the original fMRI sample rate. |
2,526 | Finite-Time Regret of Thompson Sampling Algorithms for Exponential Family Multi-Armed Bandits | We study the regret of Thompson sampling (TS) algorithms for exponential
family bandits, where the reward distribution is from a one-dimensional
exponential family, which covers many common reward distributions including
Bernoulli, Gaussian, Gamma, Exponential, etc. We propose a Thompson sampling
algorithm, termed ExpTS, which uses a novel sampling distribution to avoid the
under-estimation of the optimal arm. We provide a tight regret analysis for
ExpTS, which simultaneously yields both the finite-time regret bound as well as
the asymptotic regret bound. In particular, for a $K$-armed bandit with
exponential family rewards, ExpTS over a horizon $T$ is sub-UCB (a strong
criterion for the finite-time regret that is problem-dependent), minimax
optimal up to a factor $\sqrt{\log K}$, and asymptotically optimal, for
exponential family rewards. Moreover, we propose ExpTS$^+$, by adding a greedy
exploitation step in addition to the sampling distribution used in ExpTS, to
avoid the over-estimation of sub-optimal arms. ExpTS$^+$ is an anytime bandit
algorithm and achieves the minimax optimality and asymptotic optimality
simultaneously for exponential family reward distributions. Our proof
techniques are general and conceptually simple and can be easily applied to
analyze standard Thompson sampling with specific reward distributions. |
2,527 | How does overparametrization affect performance on minority groups? | The benefits of overparameterization for the overall performance of modern
machine learning (ML) models are well known. However, the effect of
overparameterization at a more granular level of data subgroups is less
understood. Recent empirical studies demonstrate encouraging results: (i) when
groups are not known, overparameterized models trained with empirical risk
minimization (ERM) perform better on minority groups; (ii) when groups are
known, ERM on data subsampled to equalize group sizes yields state-of-the-art
worst-group-accuracy in the overparameterized regime. In this paper, we
complement these empirical studies with a theoretical investigation of the risk
of overparameterized random feature models on minority groups. In a setting in
which the regression functions for the majority and minority groups are
different, we show that overparameterization always improves minority group
performance. |
2,528 | Few-Shot Learning by Dimensionality Reduction in Gradient Space | We introduce SubGD, a novel few-shot learning method which is based on the
recent finding that stochastic gradient descent updates tend to live in a
low-dimensional parameter subspace. In experimental and theoretical analyses,
we show that models confined to a suitable predefined subspace generalize well
for few-shot learning. A suitable subspace fulfills three criteria across the
given tasks: it (a) allows to reduce the training error by gradient flow, (b)
leads to models that generalize well, and (c) can be identified by stochastic
gradient descent. SubGD identifies these subspaces from an eigendecomposition
of the auto-correlation matrix of update directions across different tasks.
Demonstrably, we can identify low-dimensional suitable subspaces for few-shot
learning of dynamical systems, which have varying properties described by one
or few parameters of the analytical system description. Such systems are
ubiquitous among real-world applications in science and engineering. We
experimentally corroborate the advantages of SubGD on three distinct dynamical
systems problem settings, significantly outperforming popular few-shot learning
methods both in terms of sample efficiency and performance. |
2,529 | Parametric Chordal Sparsity for SDP-based Neural Network Verification | Many future technologies rely on neural networks, but verifying the
correctness of their behavior remains a major challenge. It is known that
neural networks can be fragile in the presence of even small input
perturbations, yielding unpredictable outputs. The verification of neural
networks is therefore vital to their adoption, and a number of approaches have
been proposed in recent years. In this paper we focus on semidefinite
programming (SDP) based techniques for neural network verification, which are
particularly attractive because they can encode expressive behaviors while
ensuring a polynomial time decision. Our starting point is the DeepSDP
framework proposed by Fazlyab et al, which uses quadratic constraints to
abstract the verification problem into a large-scale SDP. When the size of the
neural network grows, however, solving this SDP quickly becomes intractable.
Our key observation is that by leveraging chordal sparsity and specific
parametrizations of DeepSDP, we can decompose the primary computational
bottleneck of DeepSDP -- a large linear matrix inequality (LMI) -- into an
equivalent collection of smaller LMIs. Our parametrization admits a tunable
parameter, allowing us to trade-off efficiency and accuracy in the verification
procedure. We call our formulation Chordal-DeepSDP, and provide experimental
evaluation to show that it can: (1) effectively increase accuracy with the
tunable parameter and (2) outperform DeepSDP on deeper networks. |
2,530 | SHRED: 3D Shape Region Decomposition with Learned Local Operations | We present SHRED, a method for 3D SHape REgion Decomposition. SHRED takes a
3D point cloud as input and uses learned local operations to produce a
segmentation that approximates fine-grained part instances. We endow SHRED with
three decomposition operations: splitting regions, fixing the boundaries
between regions, and merging regions together. Modules are trained
independently and locally, allowing SHRED to generate high-quality
segmentations for categories not seen during training. We train and evaluate
SHRED with fine-grained segmentations from PartNet; using its merge-threshold
hyperparameter, we show that SHRED produces segmentations that better respect
ground-truth annotations compared with baseline methods, at any desired
decomposition granularity. Finally, we demonstrate that SHRED is useful for
downstream applications, out-performing all baselines on zero-shot fine-grained
part instance segmentation and few-shot fine-grained semantic segmentation when
combined with methods that learn to label shape regions. |
2,531 | Short Blocklength Wiretap Channel Codes via Deep Learning: Design and Performance Evaluation | We design short blocklength codes for the Gaussian wiretap channel under
information-theoretic security guarantees. Our approach consists in decoupling
the reliability and secrecy constraints in our code design. Specifically, we
handle the reliability constraint via an autoencoder, and handle the secrecy
constraint with hash functions. For blocklengths smaller than or equal to 16,
we evaluate through simulations the probability of error at the legitimate
receiver and the leakage at the eavesdropper for our code construction. This
leakage is defined as the mutual information between the confidential message
and the eavesdropper's channel observations, and is empirically measured via a
neural network-based mutual information estimator. Our simulation results
provide examples of codes with positive secrecy rates that outperform the best
known achievable secrecy rates obtained non-constructively for the Gaussian
wiretap channel. Additionally, we show that our code design is suitable for the
compound and arbitrarily varying Gaussian wiretap channels, for which the
channel statistics are not perfectly known but only known to belong to a
pre-specified uncertainty set. These models not only capture uncertainty
related to channel statistics estimation, but also scenarios where the
eavesdropper jams the legitimate transmission or influences its own channel
statistics by changing its location. |
2,532 | FDGNN: Fully Dynamic Graph Neural Network | Dynamic Graph Neural Networks recently became more and more important as
graphs from many scientific fields, ranging from mathematics, biology, social
sciences, and physics to computer science, are dynamic by nature. While
temporal changes (dynamics) play an essential role in many real-world
applications, most of the models in the literature on Graph Neural Networks
(GNN) process static graphs. The few GNN models on dynamic graphs only consider
exceptional cases of dynamics, e.g., node attribute-dynamic graphs or
structure-dynamic graphs limited to additions or changes to the graph's edges,
etc. Therefore, we present a novel Fully Dynamic Graph Neural Network (FDGNN)
that can handle fully-dynamic graphs in continuous time. The proposed method
provides a node and an edge embedding that includes their activity to address
added and deleted nodes or edges, and possible attributes. Furthermore, the
embeddings specify Temporal Point Processes for each event to encode the
distributions of the structure- and attribute-related incoming graph events. In
addition, our model can be updated efficiently by considering single events for
local retraining. |
2,533 | Discrete State-Action Abstraction via the Successor Representation | When reinforcement learning is applied with sparse rewards, agents must spend
a prohibitively long time exploring the unknown environment without any
learning signal. Abstraction is one approach that provides the agent with an
intrinsic reward for transitioning in a latent space. Prior work focuses on
dense continuous latent spaces, or requires the user to manually provide the
representation. Our approach is the first for automatically learning a discrete
abstraction of the underlying environment. Moreover, our method works on
arbitrary input spaces, using an end-to-end trainable regularized successor
representation model. For transitions between abstract states, we train a set
of temporally extended actions in the form of options, i.e., an action
abstraction. Our proposed algorithm, Discrete State-Action Abstraction (DSAA),
iteratively swaps between training these options and using them to efficiently
explore more of the environment to improve the state abstraction. As a result,
our model is not only useful for transfer learning but also in the online
learning setting. We empirically show that our agent is able to explore the
environment and solve provided tasks more efficiently than baseline
reinforcement learning algorithms. Our code is publicly available at
\url{https://github.com/amnonattali/dsaa}. |
2,534 | Adversarial Reprogramming Revisited | Adversarial reprogramming, introduced by Elsayed, Goodfellow, and
Sohl-Dickstein, seeks to repurpose a neural network to perform a different
task, by manipulating its input without modifying its weights. We prove that
two-layer ReLU neural networks with random weights can be adversarially
reprogrammed to achieve arbitrarily high accuracy on Bernoulli data models over
hypercube vertices, provided the network width is no greater than its input
dimension. We also substantially strengthen a recent result of Phuong and
Lampert on directional convergence of gradient flow, and obtain as a corollary
that training two-layer ReLU neural networks on orthogonally separable datasets
can cause their adversarial reprogramming to fail. We support these theoretical
results by experiments that demonstrate that, as long as batch normalisation
layers are suitably initialised, even untrained networks with random weights
are susceptible to adversarial reprogramming. This is in contrast to
observations in several recent works that suggested that adversarial
reprogramming is not possible for untrained networks to any degree of
reliability. |
2,535 | Combining physics-based and data-driven techniques for reliable hybrid analysis and modeling using the corrective source term approach | Upcoming technologies like digital twins, autonomous, and artificial
intelligent systems involving safety-critical applications require models which
are accurate, interpretable, computationally efficient, and generalizable.
Unfortunately, the two most commonly used modeling approaches, physics-based
modeling (PBM) and data-driven modeling (DDM) fail to satisfy all these
requirements. In the current work, we demonstrate how a hybrid approach
combining the best of PBM and DDM can result in models which can outperform
them both. We do so by combining partial differential equations based on first
principles describing partially known physics with a black box DDM, in this
case, a deep neural network model compensating for the unknown physics. First,
we present a mathematical argument for why this approach should work and then
apply the hybrid approach to model two dimensional heat diffusion problem with
an unknown source term. The result demonstrates the method's superior
performance in terms of accuracy, and generalizability. Additionally, it is
shown how the DDM part can be interpreted within the hybrid framework to make
the overall approach reliable. |
2,536 | FEL: High Capacity Learning for Recommendation and Ranking via Federated Ensemble Learning | Federated learning (FL) has emerged as an effective approach to address
consumer privacy needs. FL has been successfully applied to certain machine
learning tasks, such as training smart keyboard models and keyword spotting.
Despite FL's initial success, many important deep learning use cases, such as
ranking and recommendation tasks, have been limited from on-device learning.
One of the key challenges faced by practical FL adoption for DL-based ranking
and recommendation is the prohibitive resource requirements that cannot be
satisfied by modern mobile systems. We propose Federated Ensemble Learning
(FEL) as a solution to tackle the large memory requirement of deep learning
ranking and recommendation tasks. FEL enables large-scale ranking and
recommendation model training on-device by simultaneously training multiple
model versions on disjoint clusters of client devices. FEL integrates the
trained sub-models via an over-arch layer into an ensemble model that is hosted
on the server. Our experiments demonstrate that FEL leads to 0.43-2.31% model
quality improvement over traditional on-device federated learning - a
significant improvement for ranking and recommendation system use cases. |
2,537 | Learning in Observable POMDPs, without Computationally Intractable Oracles | Much of reinforcement learning theory is built on top of oracles that are
computationally hard to implement. Specifically for learning near-optimal
policies in Partially Observable Markov Decision Processes (POMDPs), existing
algorithms either need to make strong assumptions about the model dynamics
(e.g. deterministic transitions) or assume access to an oracle for solving a
hard optimistic planning or estimation problem as a subroutine. In this work we
develop the first oracle-free learning algorithm for POMDPs under reasonable
assumptions. Specifically, we give a quasipolynomial-time end-to-end algorithm
for learning in "observable" POMDPs, where observability is the assumption that
well-separated distributions over states induce well-separated distributions
over observations. Our techniques circumvent the more traditional approach of
using the principle of optimism under uncertainty to promote exploration, and
instead give a novel application of barycentric spanners to constructing policy
covers. |
2,538 | Robust Sparse Mean Estimation via Sum of Squares | We study the problem of high-dimensional sparse mean estimation in the
presence of an $\epsilon$-fraction of adversarial outliers. Prior work obtained
sample and computationally efficient algorithms for this task for
identity-covariance subgaussian distributions. In this work, we develop the
first efficient algorithms for robust sparse mean estimation without a priori
knowledge of the covariance. For distributions on $\mathbb R^d$ with
"certifiably bounded" $t$-th moments and sufficiently light tails, our
algorithm achieves error of $O(\epsilon^{1-1/t})$ with sample complexity $m =
(k\log(d))^{O(t)}/\epsilon^{2-2/t}$. For the special case of the Gaussian
distribution, our algorithm achieves near-optimal error of $\tilde O(\epsilon)$
with sample complexity $m = O(k^4 \mathrm{polylog}(d))/\epsilon^2$. Our
algorithms follow the Sum-of-Squares based, proofs to algorithms approach. We
complement our upper bounds with Statistical Query and low-degree polynomial
testing lower bounds, providing evidence that the sample-time-error tradeoffs
achieved by our algorithms are qualitatively the best possible. |
2,539 | A Benchmark for Federated Hetero-Task Learning | To investigate the heterogeneity in federated learning in real-world
scenarios, we generalize the classic federated learning to federated
hetero-task learning, which emphasizes the inconsistency across the
participants in federated learning in terms of both data distribution and
learning tasks. We also present B-FHTL, a federated hetero-task learning
benchmark consisting of simulation dataset, FL protocols and a unified
evaluation mechanism. B-FHTL dataset contains three well-designed federated
learning tasks with increasing heterogeneity. Each task simulates the clients
with different non-IID data and learning tasks. To ensure fair comparison among
different FL algorithms, B-FHTL builds in a full suite of FL protocols by
providing high-level APIs to avoid privacy leakage, and presets most common
evaluation metrics spanning across different learning tasks, such as
regression, classification, text generation and etc. Furthermore, we compare
the FL algorithms in fields of federated multi-task learning, federated
personalization and federated meta learning within B-FHTL, and highlight the
influence of heterogeneity and difficulties of federated hetero-task learning.
Our benchmark, including the federated dataset, protocols, the evaluation
mechanism and the preliminary experiment, is open-sourced at
https://github.com/alibaba/FederatedScope/tree/master/benchmark/B-FHTL |
2,540 | Towards Bridging Algorithm and Theory for Unbiased Recommendation | This work studies the problem of learning unbiased algorithms from biased
feedback for recommender systems. We address this problem from both theoretical
and algorithmic perspectives. Recent works in unbiased learning have advanced
the state-of-the-art with various techniques such as meta-learning, knowledge
distillation, and information bottleneck. Despite their empirical successes,
most of them lack theoretical guarantee, forming non-negligible gaps between
the theories and recent algorithms. To this end, we first view the unbiased
recommendation problem from a distribution shift perspective. We theoretically
analyze the generalization bounds of unbiased learning and suggest their close
relations with recent unbiased learning objectives. Based on the theoretical
analysis, we further propose a principled framework, Adversarial Self-Training
(AST), for unbiased recommendation. Empirical evaluation on real-world and
semi-synthetic datasets demonstrate the effectiveness of the proposed AST. |
2,541 | Generating Long Videos of Dynamic Scenes | We present a video generation model that accurately reproduces object motion,
changes in camera viewpoint, and new content that arises over time. Existing
video generation methods often fail to produce new content as a function of
time while maintaining consistencies expected in real environments, such as
plausible dynamics and object persistence. A common failure case is for content
to never change due to over-reliance on inductive biases to provide temporal
consistency, such as a single latent code that dictates content for the entire
video. On the other extreme, without long-term consistency, generated videos
may morph unrealistically between different scenes. To address these
limitations, we prioritize the time axis by redesigning the temporal latent
representation and learning long-term consistency from data by training on
longer videos. To this end, we leverage a two-phase training strategy, where we
separately train using longer videos at a low resolution and shorter videos at
a high resolution. To evaluate the capabilities of our model, we introduce two
new benchmark datasets with explicit focus on long-term temporal dynamics. |
2,542 | Improving Fairness in Graph Neural Networks via Mitigating Sensitive Attribute Leakage | Graph Neural Networks (GNNs) have shown great power in learning node
representations on graphs. However, they may inherit historical prejudices from
training data, leading to discriminatory bias in predictions. Although some
work has developed fair GNNs, most of them directly borrow fair representation
learning techniques from non-graph domains without considering the potential
problem of sensitive attribute leakage caused by feature propagation in GNNs.
However, we empirically observe that feature propagation could vary the
correlation of previously innocuous non-sensitive features to the sensitive
ones. This can be viewed as a leakage of sensitive information which could
further exacerbate discrimination in predictions. Thus, we design two feature
masking strategies according to feature correlations to highlight the
importance of considering feature propagation and correlation variation in
alleviating discrimination. Motivated by our analysis, we propose Fair View
Graph Neural Network (FairVGNN) to generate fair views of features by
automatically identifying and masking sensitive-correlated features considering
correlation variation after feature propagation. Given the learned fair views,
we adaptively clamp weights of the encoder to avoid using sensitive-related
features. Experiments on real-world datasets demonstrate that FairVGNN enjoys a
better trade-off between model utility and fairness. Our code is publicly
available at https://github.com/YuWVandy/FairVGNN. |
2,543 | Detecting Anomalous Cryptocurrency Transactions: an AML/CFT Application of Machine Learning-based Forensics | The rise of blockchain and distributed ledger technologies (DLTs) in the
financial sector has generated a socio-economic shift that triggered legal
concerns and regulatory initiatives. While the anonymity of DLTs may safeguard
the right to privacy, data protection and other civil liberties, lack of
identification hinders accountability, investigation and enforcement. The
resulting challenges extend to the rules to combat money laundering and the
financing of terrorism and proliferation (AML/CFT). As law enforcement agencies
and analytics companies have begun to successfully apply forensics to track
currency across blockchain ecosystems, in this paper we focus on the increasing
relevance of these techniques. In particular, we offer insights into the
application to the Internet of Money (IoM) of machine learning, network and
transaction graph analysis. After providing some background on the notion of
anonymity in the IoM and on the interplay between AML/CFT and blockchain
forensics, we focus on anomaly detection approaches leading to our experiments.
Namely, we analyzed a real-world dataset of Bitcoin transactions represented as
a directed graph network through various machine learning techniques. Our claim
is that the AML/CFT domain could benefit from novel graph analysis methods in
machine learning. Indeed, our findings show that the Graph Convolutional
Networks (GCN) and Graph Attention Networks (GAT) neural network types
represent a promising solution for AML/CFT compliance. |
2,544 | FedRel: An Adaptive Federated Relevance Framework for Spatial Temporal Graph Learning | Spatial-temporal data contains rich information and has been widely studied
in recent years due to the rapid development of relevant applications in many
fields. For instance, medical institutions often use electrodes attached to
different parts of a patient to analyse the electorencephal data rich with
spatial and temporal features for health assessment and disease diagnosis.
Existing research has mainly used deep learning techniques such as
convolutional neural network (CNN) or recurrent neural network (RNN) to extract
hidden spatial-temporal features. Yet, it is challenging to incorporate both
inter-dependencies spatial information and dynamic temporal changes
simultaneously. In reality, for a model that leverages these spatial-temporal
features to fulfil complex prediction tasks, it often requires a colossal
amount of training data in order to obtain satisfactory model performance.
Considering the above-mentioned challenges, we propose an adaptive federated
relevance framework, namely FedRel, for spatial-temporal graph learning in this
paper. After transforming the raw spatial-temporal data into high quality
features, the core Dynamic Inter-Intra Graph (DIIG) module in the framework is
able to use these features to generate the spatial-temporal graphs capable of
capturing the hidden topological and long-term temporal correlation information
in these graphs. To improve the model generalization ability and performance
while preserving the local data privacy, we also design a relevance-driven
federated learning module in our framework to leverage diverse data
distributions from different participants with attentive aggregations of their
models. |
2,545 | Two-stage Human Activity Recognition on Microcontrollers with Decision Trees and CNNs | Human Activity Recognition (HAR) has become an increasingly popular task for
embedded devices such as smartwatches. Most HAR systems for ultra-low power
devices are based on classic Machine Learning (ML) models, whereas Deep
Learning (DL), although reaching state-of-the-art accuracy, is less popular due
to its high energy consumption, which poses a significant challenge for
battery-operated and resource-constrained devices. In this work, we bridge the
gap between on-device HAR and DL thanks to a hierarchical architecture composed
of a decision tree (DT) and a one dimensional Convolutional Neural Network (1D
CNN). The two classifiers operate in a cascaded fashion on two different
sub-tasks: the DT classifies only the easiest activities, while the CNN deals
with more complex ones. With experiments on a state-of-the-art dataset and
targeting a single-core RISC-V MCU, we show that this approach allows to save
up to 67.7% energy w.r.t. a "stand-alone" DL architecture at iso-accuracy.
Additionally, the two-stage system either introduces a negligible memory
overhead (up to 200 B) or on the contrary, reduces the total memory occupation. |
2,546 | Towards a General Purpose CNN for Long Range Dependencies in $N$D | The use of Convolutional Neural Networks (CNNs) is widespread in Deep
Learning due to a range of desirable model properties which result in an
efficient and effective machine learning framework. However, performant CNN
architectures must be tailored to specific tasks in order to incorporate
considerations such as the input length, resolution, and dimentionality. In
this work, we overcome the need for problem-specific CNN architectures with our
Continuous Convolutional Neural Network (CCNN): a single CNN architecture
equipped with continuous convolutional kernels that can be used for tasks on
data of arbitrary resolution, dimensionality and length without structural
changes. Continuous convolutional kernels model long range dependencies at
every layer, and remove the need for downsampling layers and task-dependent
depths needed in current CNN architectures. We show the generality of our
approach by applying the same CCNN to a wide set of tasks on sequential
(1$\mathrm{D}$) and visual data (2$\mathrm{D}$). Our CCNN performs
competitively and often outperforms the current state-of-the-art across all
tasks considered. |
2,547 | Group privacy for personalized federated learning | Federated learning is a type of collaborative machine learning, where
participating clients process their data locally, sharing only updates to the
collaborative model. This enables to build privacy-aware distributed machine
learning models, among others. The goal is the optimization of a statistical
model's parameters by minimizing a cost function of a collection of datasets
which are stored locally by a set of clients. This process exposes the clients
to two issues: leakage of private information and lack of personalization of
the model. On the other hand, with the recent advancements in techniques to
analyze data, there is a surge of concern for the privacy violation of the
participating clients. To mitigate this, differential privacy and its variants
serve as a standard for providing formal privacy guarantees. Often the clients
represent very heterogeneous communities and hold data which are very diverse.
Therefore, aligned with the recent focus of the FL community to build a
framework of personalized models for the users representing their diversity, it
is also of utmost importance to protect against potential threats against the
sensitive and personal information of the clients. $d$-privacy, which is a
generalization of geo-indistinguishability, the lately popularized paradigm of
location privacy, uses a metric-based obfuscation technique that preserves the
spatial distribution of the original data. To address the issue of protecting
the privacy of the clients and allowing for personalized model training to
enhance the fairness and utility of the system, we propose a method to provide
group privacy guarantees exploiting some key properties of $d$-privacy which
enables personalized models under the framework of FL. We provide with
theoretical justifications to the applicability and experimental validation on
real-world datasets to illustrate the working of the proposed method. |
2,548 | Towards Understanding and Mitigating Audio Adversarial Examples for Speaker Recognition | Speaker recognition systems (SRSs) have recently been shown to be vulnerable
to adversarial attacks, raising significant security concerns. In this work, we
systematically investigate transformation and adversarial training based
defenses for securing SRSs. According to the characteristic of SRSs, we present
22 diverse transformations and thoroughly evaluate them using 7 recent
promising adversarial attacks (4 white-box and 3 black-box) on speaker
recognition. With careful regard for best practices in defense evaluations, we
analyze the strength of transformations to withstand adaptive attacks. We also
evaluate and understand their effectiveness against adaptive attacks when
combined with adversarial training. Our study provides lots of useful insights
and findings, many of them are new or inconsistent with the conclusions in the
image and speech recognition domains, e.g., variable and constant bit rate
speech compressions have different performance, and some non-differentiable
transformations remain effective against current promising evasion techniques
which often work well in the image domain. We demonstrate that the proposed
novel feature-level transformation combined with adversarial training is rather
effective compared to the sole adversarial training in a complete white-box
setting, e.g., increasing the accuracy by 13.62% and attack cost by two orders
of magnitude, while other transformations do not necessarily improve the
overall defense capability. This work sheds further light on the research
directions in this field. We also release our evaluation platform SPEAKERGUARD
to foster further research. |
2,549 | Data Stealing Attack on Medical Images: Is it Safe to Export Networks from Data Lakes? | In privacy-preserving machine learning, it is common that the owner of the
learned model does not have any physical access to the data. Instead, only a
secured remote access to a data lake is granted to the model owner without any
ability to retrieve data from the data lake. Yet, the model owner may want to
export the trained model periodically from the remote repository and a question
arises whether this may cause is a risk of data leakage. In this paper, we
introduce the concept of data stealing attack during the export of neural
networks. It consists in hiding some information in the exported network that
allows the reconstruction outside the data lake of images initially stored in
that data lake. More precisely, we show that it is possible to train a network
that can perform lossy image compression and at the same time solve some
utility tasks such as image segmentation. The attack then proceeds by exporting
the compression decoder network together with some image codes that leads to
the image reconstruction outside the data lake. We explore the feasibility of
such attacks on databases of CT and MR images, showing that it is possible to
obtain perceptually meaningful reconstructions of the target dataset, and that
the stolen dataset can be used in turns to solve a broad range of tasks.
Comprehensive experiments and analyses show that data stealing attacks should
be considered as a threat for sensitive imaging data sources. |
2,550 | Gender Bias in Word Embeddings: A Comprehensive Analysis of Frequency, Syntax, and Semantics | The statistical regularities in language corpora encode well-known social
biases into word embeddings. Here, we focus on gender to provide a
comprehensive analysis of group-based biases in widely-used static English word
embeddings trained on internet corpora (GloVe 2014, fastText 2017). Using the
Single-Category Word Embedding Association Test, we demonstrate the widespread
prevalence of gender biases that also show differences in: (1) frequencies of
words associated with men versus women; (b) part-of-speech tags in
gender-associated words; (c) semantic categories in gender-associated words;
and (d) valence, arousal, and dominance in gender-associated words.
First, in terms of word frequency: we find that, of the 1,000 most frequent
words in the vocabulary, 77% are more associated with men than women, providing
direct evidence of a masculine default in the everyday language of the
English-speaking world. Second, turning to parts-of-speech: the top
male-associated words are typically verbs (e.g., fight, overpower) while the
top female-associated words are typically adjectives and adverbs (e.g., giving,
emotionally). Gender biases in embeddings also permeate parts-of-speech. Third,
for semantic categories: bottom-up, cluster analyses of the top 1,000 words
associated with each gender. The top male-associated concepts include roles and
domains of big tech, engineering, religion, sports, and violence; in contrast,
the top female-associated concepts are less focused on roles, including,
instead, female-specific slurs and sexual content, as well as appearance and
kitchen terms. Fourth, using human ratings of word valence, arousal, and
dominance from a ~20,000 word lexicon, we find that male-associated words are
higher on arousal and dominance, while female-associated words are higher on
valence. |
2,551 | On the Role of Discount Factor in Offline Reinforcement Learning | Offline reinforcement learning (RL) enables effective learning from
previously collected data without exploration, which shows great promise in
real-world applications when exploration is expensive or even infeasible. The
discount factor, $\gamma$, plays a vital role in improving online RL sample
efficiency and estimation accuracy, but the role of the discount factor in
offline RL is not well explored. This paper examines two distinct effects of
$\gamma$ in offline RL with theoretical analysis, namely the regularization
effect and the pessimism effect. On the one hand, $\gamma$ is a regulator to
trade-off optimality with sample efficiency upon existing offline techniques.
On the other hand, lower guidance $\gamma$ can also be seen as a way of
pessimism where we optimize the policy's performance in the worst possible
models. We empirically verify the above theoretical observation with tabular
MDPs and standard D4RL tasks. The results show that the discount factor plays
an essential role in the performance of offline RL algorithms, both under small
data regimes upon existing offline methods and in large data regimes without
other conservative methods. |
2,552 | Imitating Past Successes can be Very Suboptimal | Prior work has proposed a simple strategy for reinforcement learning (RL):
label experience with the outcomes achieved in that experience, and then
imitate the relabeled experience. These outcome-conditioned imitation learning
methods are appealing because of their simplicity, strong performance, and
close ties with supervised learning. However, it remains unclear how these
methods relate to the standard RL objective, reward maximization. In this
paper, we prove that existing outcome-conditioned imitation learning methods do
not necessarily improve the policy; rather, in some settings they can decrease
the expected reward. Nonetheless, we show that a simple modification results in
a method that does guarantee policy improvement, under some assumptions. Our
aim is not to develop an entirely new method, but rather to explain how a
variant of outcome-conditioned imitation learning can be used to maximize
rewards. |
2,553 | On Outer Bi-Lipschitz Extensions of Linear Johnson-Lindenstrauss Embeddings of Low-Dimensional Submanifolds of $\mathbb{R}^N$ | Let $\mathcal{M}$ be a compact $d$-dimensional submanifold of $\mathbb{R}^N$
with reach $\tau$ and volume $V_{\mathcal M}$. Fix $\epsilon \in (0,1)$. In
this paper we prove that a nonlinear function $f: \mathbb{R}^N \rightarrow
\mathbb{R}^{m}$ exists with $m \leq C \left(d / \epsilon^2 \right) \log
\left(\frac{\sqrt[d]{V_{\mathcal M}}}{\tau} \right)$ such that $$(1 - \epsilon)
\| {\bf x} - {\bf y} \|_2 \leq \left\| f({\bf x}) - f({\bf y}) \right\|_2 \leq
(1 + \epsilon) \| {\bf x} - {\bf y} \|_2$$ holds for all ${\bf x} \in
\mathcal{M}$ and ${\bf y} \in \mathbb{R}^N$. In effect, $f$ not only serves as
a bi-Lipschitz function from $\mathcal{M}$ into $\mathbb{R}^{m}$ with
bi-Lipschitz constants close to one, but also approximately preserves all
distances from points not in $\mathcal{M}$ to all points in $\mathcal{M}$ in
its image. Furthermore, the proof is constructive and yields an algorithm which
works well in practice. In particular, it is empirically demonstrated herein
that such nonlinear functions allow for more accurate compressive nearest
neighbor classification than standard linear Johnson-Lindenstrauss embeddings
do in practice. |
2,554 | Computational Doob's $h$-transforms for Online Filtering of Discretely Observed Diffusions | This paper is concerned with online filtering of discretely observed
nonlinear diffusion processes. Our approach is based on the fully adapted
auxiliary particle filter, which involves Doob's $h$-transforms that are
typically intractable. We propose a computational framework to approximate
these $h$-transforms by solving the underlying backward Kolmogorov equations
using nonlinear Feynman-Kac formulas and neural networks. The methodology
allows one to train a locally optimal particle filter prior to the
data-assimilation procedure. Numerical experiments illustrate that the proposed
approach can be orders of magnitude more efficient than the bootstrap particle
filter in the regime of highly informative observations, when the observations
are extreme under the model, and if the state dimension is large. |
2,555 | DeepOPF-AL: Augmented Learning for Solving AC-OPF Problems with Multiple Load-Solution Mappings | The existence of multiple load-solution mappings of non-convex AC-OPF
problems poses a fundamental challenge to deep neural network (DNN) schemes. As
the training dataset may contain a mixture of data points corresponding to
different load-solution mappings, the DNN can fail to learn a legitimate
mapping and generate inferior solutions. We propose DeepOPF-AL as an
augmented-learning approach to tackle this issue. The idea is to train a DNN to
learn a unique mapping from an augmented input, i.e., (load, initial point), to
the solution generated by an iterative OPF solver with the load and initial
point as intake. We then apply the learned augmented mapping to solve AC-OPF
problems much faster than conventional solvers. Simulation results over IEEE
test cases show that DeepOPF-AL achieves noticeably better optimality and
similar feasibility and speedup performance, as compared to a recent DNN
scheme, with the same DNN size yet elevated training complexity. |
2,556 | Building Robust Ensembles via Margin Boosting | In the context of adversarial robustness, a single model does not usually
have enough power to defend against all possible adversarial attacks, and as a
result, has sub-optimal robustness. Consequently, an emerging line of work has
focused on learning an ensemble of neural networks to defend against
adversarial attacks. In this work, we take a principled approach towards
building robust ensembles. We view this problem from the perspective of
margin-boosting and develop an algorithm for learning an ensemble with maximum
margin. Through extensive empirical evaluation on benchmark datasets, we show
that our algorithm not only outperforms existing ensembling techniques, but
also large models trained in an end-to-end fashion. An important byproduct of
our work is a margin-maximizing cross-entropy (MCE) loss, which is a better
alternative to the standard cross-entropy (CE) loss. Empirically, we show that
replacing the CE loss in state-of-the-art adversarial training techniques with
our MCE loss leads to significant performance improvement. |
2,557 | An efficient semi-supervised quality control system trained using physics-based MRI-artefact generators and adversarial training | Large medical imaging data sets are becoming increasingly available. A common
challenge in these data sets is to ensure that each sample meets minimum
quality requirements devoid of significant artefacts. Despite a wide range of
existing automatic methods having been developed to identify imperfections and
artefacts in medical imaging, they mostly rely on data-hungry methods. In
particular, the lack of sufficient scans with artefacts available for training
has created a barrier in designing and deploying machine learning in clinical
research. To tackle this problem, we propose a novel framework having four main
components: (1) a set of artefact generators inspired by magnetic resonance
physics to corrupt brain MRI scans and augment a training dataset, (2) a set of
abstract and engineered features to represent images compactly, (3) a feature
selection process that depends on the class of artefact to improve
classification performance, and (4) a set of Support Vector Machine (SVM)
classifiers trained to identify artefacts. Our novel contributions are
threefold: first, we use the novel physics-based artefact generators to
generate synthetic brain MRI scans with controlled artefacts as a data
augmentation technique. This will avoid the labour-intensive collection and
labelling process of scans with rare artefacts. Second, we propose a large pool
of abstract and engineered image features developed to identify 9 different
artefacts for structural MRI. Finally, we use an artefact-based feature
selection block that, for each class of artefacts, finds the set of features
that provide the best classification performance. We performed validation
experiments on a large data set of scans with artificially-generated artefacts,
and in a multiple sclerosis clinical trial where real artefacts were identified
by experts, showing that the proposed pipeline outperforms traditional methods. |
2,558 | Adaptive Regularization for Adversarial Training | Adversarial training, which is to enhance robustness against adversarial
attacks, has received much attention because it is easy to generate
human-imperceptible perturbations of data to deceive a given deep neural
network. In this paper, we propose a new adversarial training algorithm that is
theoretically well motivated and empirically superior to other existing
algorithms. A novel feature of the proposed algorithm is to use a data-adaptive
regularization for robustifying a prediction model. We apply more
regularization to data which are more vulnerable to adversarial attacks and
vice versa. Even though the idea of data-adaptive regularization is not new,
our data-adaptive regularization has a firm theoretical base of reducing an
upper bound of the robust risk. Numerical experiments illustrate that our
proposed algorithm improves the generalization (accuracy on clean samples) and
robustness (accuracy on adversarial attacks) simultaneously to achieve the
state-of-the-art performance. |
2,559 | AS2T: Arbitrary Source-To-Target Adversarial Attack on Speaker Recognition Systems | Recent work has illuminated the vulnerability of speaker recognition systems
(SRSs) against adversarial attacks, raising significant security concerns in
deploying SRSs. However, they considered only a few settings (e.g., some
combinations of source and target speakers), leaving many interesting and
important settings in real-world attack scenarios alone. In this work, we
present AS2T, the first attack in this domain which covers all the settings,
thus allows the adversary to craft adversarial voices using arbitrary source
and target speakers for any of three main recognition tasks. Since none of the
existing loss functions can be applied to all the settings, we explore many
candidate loss functions for each setting including the existing and newly
designed ones. We thoroughly evaluate their efficacy and find that some
existing loss functions are suboptimal. Then, to improve the robustness of AS2T
towards practical over-the-air attack, we study the possible distortions
occurred in over-the-air transmission, utilize different transformation
functions with different parameters to model those distortions, and incorporate
them into the generation of adversarial voices. Our simulated over-the-air
evaluation validates the effectiveness of our solution in producing robust
adversarial voices which remain effective under various hardware devices and
various acoustic environments with different reverberation, ambient noises, and
noise levels. Finally, we leverage AS2T to perform thus far the largest-scale
evaluation to understand transferability among 14 diverse SRSs. The
transferability analysis provides many interesting and useful insights which
challenge several findings and conclusion drawn in previous works in the image
domain. Our study also sheds light on future directions of adversarial attacks
in the speaker recognition domain. |
2,560 | Improving trajectory calculations using deep learning inspired single image superresolution | Lagrangian trajectory or particle dispersion models as well as
semi-Lagrangian advection schemes require meteorological data such as wind,
temperature and geopotential at the exact spatio-temporal locations of the
particles that move independently from a regular grid. Traditionally, this
high-resolution data has been obtained by interpolating the meteorological
parameters from the gridded data of a meteorological model or reanalysis, e.g.
using linear interpolation in space and time. However, interpolation errors are
a large source of error for these models. Reducing them requires meteorological
input fields with high space and time resolution, which may not always be
available and can cause severe data storage and transfer problems. Here, we
interpret this problem as a single image superresolution task. We interpret
meteorological fields available at their native resolution as low-resolution
images and train deep neural networks to up-scale them to higher resolution,
thereby providing more accurate data for Lagrangian models. We train various
versions of the state-of-the-art Enhanced Deep Residual Networks for
Superresolution on low-resolution ERA5 reanalysis data with the goal to
up-scale these data to arbitrary spatial resolution. We show that the resulting
up-scaled wind fields have root-mean-squared errors half the size of the winds
obtained with linear spatial interpolation at acceptable computational
inference costs. In a test setup using the Lagrangian particle dispersion model
FLEXPART and reduced-resolution wind fields, we demonstrate that absolute
horizontal transport deviations of calculated trajectories from "ground-truth"
trajectories calculated with undegraded 0.5{\deg} winds are reduced by at least
49.5% (21.8%) after 48 hours relative to trajectories using linear
interpolation of the wind data when training on 2{\deg} to 1{\deg} (4{\deg} to
2{\deg}) resolution data. |
2,561 | Preconditioned Gradient Descent for Overparameterized Nonconvex Burer--Monteiro Factorization with Global Optimality Certification | We consider using gradient descent to minimize the nonconvex function
$f(X)=\phi(XX^{T})$ over an $n\times r$ factor matrix $X$, in which $\phi$ is
an underlying smooth convex cost function defined over $n\times n$ matrices.
While only a second-order stationary point $X$ can be provably found in
reasonable time, if $X$ is additionally rank deficient, then its rank
deficiency certifies it as being globally optimal. This way of certifying
global optimality necessarily requires the search rank $r$ of the current
iterate $X$ to be overparameterized with respect to the rank $r^{\star}$ of the
global minimizer $X^{\star}$. Unfortunately, overparameterization significantly
slows down the convergence of gradient descent, from a linear rate with
$r=r^{\star}$ to a sublinear rate when $r>r^{\star}$, even when $\phi$ is
strongly convex. In this paper, we propose an inexpensive preconditioner that
restores the convergence rate of gradient descent back to linear in the
overparameterized case, while also making it agnostic to possible
ill-conditioning in the global minimizer $X^{\star}$. |
2,562 | Parotid Gland MRI Segmentation Based on Swin-Unet and Multimodal Images | Parotid gland tumors account for approximately 2% to 10% of head and neck
tumors. Preoperative tumor localization, differential diagnosis, and subsequent
selection of appropriate treatment for parotid gland tumors is critical.
However, the relative rarity of these tumors and the highly dispersed tissue
types have left an unmet need for a subtle differential diagnosis of such
neoplastic lesions based on preoperative radiomics. Recently, deep learning
methods have developed rapidly, especially Transformer beats the traditional
convolutional neural network in computer vision. Many new Transformer-based
networks have been proposed for computer vision tasks. In this study,
multicenter multimodal parotid gland MRI images were collected. The Swin-Unet
which was based on Transformer was used. MRI images of STIR, T1 and T2
modalities were combined into a three-channel data to train the network. We
achieved segmentation of the region of interest for parotid gland and tumor.
The DSC of the model on the test set was 88.63%, MPA was 99.31%, MIoU was
83.99%, and HD was 3.04. Then a series of comparison experiments were designed
in this paper to further validate the segmentation performance of the
algorithm. |
2,563 | Assessing Project-Level Fine-Tuning of ML4SE Models | Machine Learning for Software Engineering (ML4SE) is an actively growing
research area that focuses on methods that help programmers in their work. In
order to apply the developed methods in practice, they need to achieve
reasonable quality in order to help rather than distract developers. While the
development of new approaches to code representation and data collection
improves the overall quality of the models, it does not take into account the
information that we can get from the project at hand.
In this work, we investigate how the model's quality can be improved if we
target a specific project. We develop a framework to assess quality
improvements that models can get after fine-tuning for the method name
prediction task on a particular project. We evaluate three models of different
complexity and compare their quality in three settings: trained on a large
dataset of Java projects, further fine-tuned on the data from a particular
project, and trained from scratch on this data. We show that per-project
fine-tuning can greatly improve the models' quality as they capture the
project's domain and naming conventions. We open-source the tool we used for
data collection, as well as the code to run the experiments:
https://zenodo.org/record/6040745. |
2,564 | Improving the Diagnosis of Psychiatric Disorders with Self-Supervised Graph State Space Models | Single subject prediction of brain disorders from neuroimaging data has
gained increasing attention in recent years. Yet, for some heterogeneous
disorders such as major depression disorder (MDD) and autism spectrum disorder
(ASD), the performance of prediction models on large-scale multi-site datasets
remains poor. We present a two-stage framework to improve the diagnosis of
heterogeneous psychiatric disorders from resting-state functional magnetic
resonance imaging (rs-fMRI). First, we propose a self-supervised mask
prediction task on data from healthy individuals that can exploit differences
between healthy controls and patients in clinical datasets. Next, we train a
supervised classifier on the learned discriminative representations. To model
rs-fMRI data, we develop Graph-S4; an extension to the recently proposed
state-space model S4 to graph settings where the underlying graph structure is
not known in advance. We show that combining the framework and Graph-S4 can
significantly improve the diagnostic performance of neuroimaging-based single
subject prediction models of MDD and ASD on three open-source multi-center
rs-fMRI clinical datasets. |
2,565 | DeepTPI: Test Point Insertion with Deep Reinforcement Learning | Test point insertion (TPI) is a widely used technique for testability
enhancement, especially for logic built-in self-test (LBIST) due to its
relatively low fault coverage. In this paper, we propose a novel TPI approach
based on deep reinforcement learning (DRL), named DeepTPI. Unlike previous
learning-based solutions that formulate the TPI task as a supervised-learning
problem, we train a novel DRL agent, instantiated as the combination of a graph
neural network (GNN) and a Deep Q-Learning network (DQN), to maximize the test
coverage improvement. Specifically, we model circuits as directed graphs and
design a graph-based value network to estimate the action values for inserting
different test points. The policy of the DRL agent is defined as selecting the
action with the maximum value. Moreover, we apply the general node embeddings
from a pre-trained model to enhance node features, and propose a dedicated
testability-aware attention mechanism for the value network. Experimental
results on circuits with various scales show that DeepTPI significantly
improves test coverage compared to the commercial DFT tool. The code of this
work is available at https://github.com/cure-lab/DeepTPI. |
2,566 | Concentration bounds for SSP Q-learning for average cost MDPs | We derive a concentration bound for a Q-learning algorithm for average cost
Markov decision processes based on an equivalent shortest path problem, and
compare it numerically with the alternative scheme based on relative value
iteration. |
2,567 | Subject Membership Inference Attacks in Federated Learning | Privacy in Federated Learning (FL) is studied at two different granularities:
item-level, which protects individual data points, and user-level, which
protects each user (participant) in the federation. Nearly all of the private
FL literature is dedicated to studying privacy attacks and defenses at these
two granularities. Recently, subject-level privacy has emerged as an
alternative privacy granularity to protect the privacy of individuals (data
subjects) whose data is spread across multiple (organizational) users in
cross-silo FL settings. An adversary might be interested in recovering private
information about these individuals (a.k.a. \emph{data subjects}) by attacking
the trained model. A systematic study of these patterns requires complete
control over the federation, which is impossible with real-world datasets. We
design a simulator for generating various synthetic federation configurations,
enabling us to study how properties of the data, model design and training, and
the federation itself impact subject privacy risk. We propose three attacks for
\emph{subject membership inference} and examine the interplay between all
factors within a federation that affect the attacks' efficacy. We also
investigate the effectiveness of Differential Privacy in mitigating this
threat. Our takeaways generalize to real-world datasets like FEMNIST, giving
credence to our findings. |
2,568 | Neural Network Decoders for Permutation Codes Correcting Different Errors | Permutation codes were extensively studied in order to correct different
types of errors for the applications on power line communication and rank
modulation for flash memory. In this paper, we introduce the neural network
decoders for permutation codes to correct these errors with one-shot decoding,
which treat the decoding as $n$ classification tasks for non-binary symbols for
a code of length $n$. These are actually the first general decoders introduced
to deal with any error type for these two applications. The performance of the
decoders is evaluated by simulations with different error models. |
2,569 | Integrating Random Effects in Deep Neural Networks | Modern approaches to supervised learning like deep neural networks (DNNs)
typically implicitly assume that observed responses are statistically
independent. In contrast, correlated data are prevalent in real-life
large-scale applications, with typical sources of correlation including
spatial, temporal and clustering structures. These correlations are either
ignored by DNNs, or ad-hoc solutions are developed for specific use cases. We
propose to use the mixed models framework to handle correlated data in DNNs. By
treating the effects underlying the correlation structure as random effects,
mixed models are able to avoid overfitted parameter estimates and ultimately
yield better predictive performance. The key to combining mixed models and DNNs
is using the Gaussian negative log-likelihood (NLL) as a natural loss function
that is minimized with DNN machinery including stochastic gradient descent
(SGD). Since NLL does not decompose like standard DNN loss functions, the use
of SGD with NLL presents some theoretical and implementation challenges, which
we address. Our approach which we call LMMNN is demonstrated to improve
performance over natural competitors in various correlation scenarios on
diverse simulated and real datasets. Our focus is on a regression setting and
tabular datasets, but we also show some results for classification. Our code is
available at https://github.com/gsimchoni/lmmnn. |
2,570 | PyTSK: A Python Toolbox for TSK Fuzzy Systems | This paper presents PyTSK, a Python toolbox for developing Takagi-Sugeno-Kang
(TSK) fuzzy systems. Based on scikit-learn and PyTorch, PyTSK allows users to
optimize TSK fuzzy systems using fuzzy clustering or mini-batch gradient
descent (MBGD) based algorithms. Several state-of-the-art MBGD-based
optimization algorithms are implemented in the toolbox, which can improve the
generalization performance of TSK fuzzy systems, especially for big data
applications. PyTSK can also be easily extended and customized for more
complicated algorithms, such as modifying the structure of TSK fuzzy systems,
developing more sophisticated training algorithms, and combining TSK fuzzy
systems with neural networks. The code of PyTSK can be found at
https://github.com/YuqiCui/pytsk. |
2,571 | Physics-Inspired Temporal Learning of Quadrotor Dynamics for Accurate Model Predictive Trajectory Tracking | Accurately modeling quadrotor's system dynamics is critical for guaranteeing
agile, safe, and stable navigation. The model needs to capture the system
behavior in multiple flight regimes and operating conditions, including those
producing highly nonlinear effects such as aerodynamic forces and torques,
rotor interactions, or possible system configuration modifications. Classical
approaches rely on handcrafted models and struggle to generalize and scale to
capture these effects. In this paper, we present a novel Physics-Inspired
Temporal Convolutional Network (PI-TCN) approach to learning quadrotor's system
dynamics purely from robot experience. Our approach combines the expressive
power of sparse temporal convolutions and dense feed-forward connections to
make accurate system predictions. In addition, physics constraints are embedded
in the training process to facilitate the network's generalization capabilities
to data outside the training distribution. Finally, we design a model
predictive control approach that incorporates the learned dynamics for accurate
closed-loop trajectory tracking fully exploiting the learned model predictions
in a receding horizon fashion. Experimental results demonstrate that our
approach accurately extracts the structure of the quadrotor's dynamics from
data, capturing effects that would remain hidden to classical approaches. To
the best of our knowledge, this is the first time physics-inspired deep
learning is successfully applied to temporal convolutional networks and to the
system identification task, while concurrently enabling predictive control. |
2,572 | On the balance between the training time and interpretability of neural ODE for time series modelling | Most machine learning methods are used as a black box for modelling. We may
try to extract some knowledge from physics-based training methods, such as
neural ODE (ordinary differential equation). Neural ODE has advantages like a
possibly higher class of represented functions, the extended interpretability
compared to black-box machine learning models, ability to describe both trend
and local behaviour. Such advantages are especially critical for time series
with complicated trends. However, the known drawback is the high training time
compared to the autoregressive models and long-short term memory (LSTM)
networks widely used for data-driven time series modelling. Therefore, we
should be able to balance interpretability and training time to apply neural
ODE in practice. The paper shows that modern neural ODE cannot be reduced to
simpler models for time-series modelling applications. The complexity of neural
ODE is compared to or exceeds the conventional time-series modelling tools. The
only interpretation that could be extracted is the eigenspace of the operator,
which is an ill-posed problem for a large system. Spectra could be extracted
using different classical analysis methods that do not have the drawback of
extended time. Consequently, we reduce the neural ODE to a simpler linear form
and propose a new view on time-series modelling using combined neural networks
and an ODE system approach. |
2,573 | Recent Advances in Bayesian Optimization | Bayesian optimization has emerged at the forefront of expensive black-box
optimization due to its data efficiency. Recent years have witnessed a
proliferation of studies on the development of new Bayesian optimization
algorithms and their applications. Hence, this paper attempts to provide a
comprehensive and updated survey of recent advances in Bayesian optimization
and identify interesting open problems. We categorize the existing work on
Bayesian optimization into nine main groups according to the motivations and
focus of the proposed algorithms. For each category, we present the main
advances with respect to the construction of surrogate models and adaptation of
the acquisition functions. Finally, we discuss the open questions and suggest
promising future research directions, in particular with regard to
heterogeneity, privacy preservation, and fairness in distributed and federated
optimization systems. |
2,574 | Generalization Error Bounds for Deep Neural Networks Trained by SGD | Generalization error bounds for deep neural networks trained by stochastic
gradient descent (SGD) are derived by combining a dynamical control of an
appropriate parameter norm and the Rademacher complexity estimate based on
parameter norms. The bounds explicitly depend on the loss along the training
trajectory, and work for a wide range of network architectures including
multilayer perceptron (MLP) and convolutional neural networks (CNN). Compared
with other algorithm-depending generalization estimates such as uniform
stability-based bounds, our bounds do not require $L$-smoothness of the
nonconvex loss function, and apply directly to SGD instead of Stochastic
Langevin gradient descent (SGLD). Numerical results show that our bounds are
non-vacuous and robust with the change of optimizer and network
hyperparameters. |
2,575 | Joint Manifold Learning and Density Estimation Using Normalizing Flows | Based on the manifold hypothesis, real-world data often lie on a
low-dimensional manifold, while normalizing flows as a likelihood-based
generative model are incapable of finding this manifold due to their structural
constraints. So, one interesting question arises: $\textit{"Can we find
sub-manifold(s) of data in normalizing flows and estimate the density of the
data on the sub-manifold(s)?"}$. In this paper, we introduce two approaches,
namely per-pixel penalized log-likelihood and hierarchical training, to answer
the mentioned question. We propose a single-step method for joint manifold
learning and density estimation by disentangling the transformed space obtained
by normalizing flows to manifold and off-manifold parts. This is done by a
per-pixel penalized likelihood function for learning a sub-manifold of the
data. Normalizing flows assume the transformed data is Gaussianizationed, but
this imposed assumption is not necessarily true, especially in high dimensions.
To tackle this problem, a hierarchical training approach is employed to improve
the density estimation on the sub-manifold. The results validate the
superiority of the proposed methods in simultaneous manifold learning and
density estimation using normalizing flows in terms of generated image quality
and likelihood. |
2,576 | Collaborative Intelligence Orchestration: Inconsistency-Based Fusion of Semi-Supervised Learning and Active Learning | While annotating decent amounts of data to satisfy sophisticated learning
models can be cost-prohibitive for many real-world applications. Active
learning (AL) and semi-supervised learning (SSL) are two effective, but often
isolated, means to alleviate the data-hungry problem. Some recent studies
explored the potential of combining AL and SSL to better probe the unlabeled
data. However, almost all these contemporary SSL-AL works use a simple
combination strategy, ignoring SSL and AL's inherent relation. Further, other
methods suffer from high computational costs when dealing with large-scale,
high-dimensional datasets. Motivated by the industry practice of labeling data,
we propose an innovative Inconsistency-based virtual aDvErsarial Active
Learning (IDEAL) algorithm to further investigate SSL-AL's potential
superiority and achieve mutual enhancement of AL and SSL, i.e., SSL propagates
label information to unlabeled samples and provides smoothed embeddings for AL,
while AL excludes samples with inconsistent predictions and considerable
uncertainty for SSL. We estimate unlabeled samples' inconsistency by
augmentation strategies of different granularities, including fine-grained
continuous perturbation exploration and coarse-grained data transformations.
Extensive experiments, in both text and image domains, validate the
effectiveness of the proposed algorithm, comparing it against state-of-the-art
baselines. Two real-world case studies visualize the practical industrial value
of applying and deploying the proposed data sampling algorithm. |
2,577 | On the Effectiveness of Fine-tuning Versus Meta-reinforcement Learning | Intelligent agents should have the ability to leverage knowledge from
previously learned tasks in order to learn new ones quickly and efficiently.
Meta-learning approaches have emerged as a popular solution to achieve this.
However, meta-reinforcement learning (meta-RL) algorithms have thus far been
restricted to simple environments with narrow task distributions. Moreover, the
paradigm of pretraining followed by fine-tuning to adapt to new tasks has
emerged as a simple yet effective solution in supervised and self-supervised
learning. This calls into question the benefits of meta-learning approaches
also in reinforcement learning, which typically come at the cost of high
complexity. We hence investigate meta-RL approaches in a variety of
vision-based benchmarks, including Procgen, RLBench, and Atari, where
evaluations are made on completely novel tasks. Our findings show that when
meta-learning approaches are evaluated on different tasks (rather than
different variations of the same task), multi-task pretraining with fine-tuning
on new tasks performs equally as well, or better, than meta-pretraining with
meta test-time adaptation. This is encouraging for future research, as
multi-task pretraining tends to be simpler and computationally cheaper than
meta-RL. From these findings, we advocate for evaluating future meta-RL methods
on more challenging tasks and including multi-task pretraining with fine-tuning
as a simple, yet strong baseline. |
2,578 | Machine Learning Sensors | Machine learning sensors represent a paradigm shift for the future of
embedded machine learning applications. Current instantiations of embedded
machine learning (ML) suffer from complex integration, lack of modularity, and
privacy and security concerns from data movement. This article proposes a more
data-centric paradigm for embedding sensor intelligence on edge devices to
combat these challenges. Our vision for "sensor 2.0" entails segregating sensor
input data and ML processing from the wider system at the hardware level and
providing a thin interface that mimics traditional sensors in functionality.
This separation leads to a modular and easy-to-use ML sensor device. We discuss
challenges presented by the standard approach of building ML processing into
the software stack of the controlling microprocessor on an embedded system and
how the modularity of ML sensors alleviates these problems. ML sensors increase
privacy and accuracy while making it easier for system builders to integrate ML
into their products as a simple component. We provide examples of prospective
ML sensors and an illustrative datasheet as a demonstration and hope that this
will build a dialogue to progress us towards sensor 2.0. |
2,579 | Marvolo: Programmatic Data Augmentation for Practical ML-Driven Malware Detection | Data augmentation has been rare in the cyber security domain due to technical
difficulties in altering data in a manner that is semantically consistent with
the original data. This shortfall is particularly onerous given the unique
difficulty of acquiring benign and malicious training data that runs into
copyright restrictions, and that institutions like banks and governments
receive targeted malware that will never exist in large quantities. We present
MARVOLO, a binary mutator that programmatically grows malware (and benign)
datasets in a manner that boosts the accuracy of ML-driven malware detectors.
MARVOLO employs semantics-preserving code transformations that mimic the
alterations that malware authors and defensive benign developers routinely make
in practice , allowing us to generate meaningful augmented data. Crucially,
semantics-preserving transformations also enable MARVOLO to safely propagate
labels from original to newly-generated data samples without mandating
expensive reverse engineering of binaries. Further, MARVOLO embeds several key
optimizations that keep costs low for practitioners by maximizing the density
of diverse data samples generated within a given time (or resource) budget.
Experiments using wide-ranging commercial malware datasets and a recent
ML-driven malware detector show that MARVOLO boosts accuracies by up to 5%,
while operating on only a small fraction (15%) of the potential input binaries. |
2,580 | DeepCAVE: An Interactive Analysis Tool for Automated Machine Learning | Automated Machine Learning (AutoML) is used more than ever before to support
users in determining efficient hyperparameters, neural architectures, or even
full machine learning pipelines. However, users tend to mistrust the
optimization process and its results due to a lack of transparency, making
manual tuning still widespread. We introduce DeepCAVE, an interactive framework
to analyze and monitor state-of-the-art optimization procedures for AutoML
easily and ad hoc. By aiming for full and accessible transparency, DeepCAVE
builds a bridge between users and AutoML and contributes to establishing trust.
Our framework's modular and easy-to-extend nature provides users with
automatically generated text, tables, and graphic visualizations. We show the
value of DeepCAVE in an exemplary use-case of outlier detection, in which our
framework makes it easy to identify problems, compare multiple runs and
interpret optimization processes. The package is freely available on GitHub
https://github.com/automl/DeepCAVE. |
2,581 | Decentralized Online Regularized Learning Over Random Time-Varying Graphs | We study the decentralized online regularized linear regression algorithm
over random time-varying graphs. At each time step, every node runs an online
estimation algorithm consisting of an innovation term processing its own new
measurement, a consensus term taking a weighted sum of estimations of its own
and its neighbors with additive and multiplicative communication noises and a
regularization term preventing over-fitting. It is not required that the
regression matrices and graphs satisfy special statistical assumptions such as
mutual independence, spatio-temporal independence or stationarity. We develop
the nonnegative supermartingale inequality of the estimation error, and prove
that the estimations of all nodes converge to the unknown true parameter vector
almost surely if the algorithm gains, graphs and regression matrices jointly
satisfy the sample path spatio-temporal persistence of excitation condition.
Especially, this condition holds by choosing appropriate algorithm gains if the
graphs are uniformly conditionally jointly connected and conditionally
balanced, and the regression models of all nodes are uniformly conditionally
spatio-temporally jointly observable, under which the algorithm converges in
mean square and almost surely. In addition, we prove that the regret upper
bound $\mathcal O(T^{1-\tau}\ln T)$, where $\tau\in (0.5,1)$ is a constant
depending on the algorithm gains. |
2,582 | Analyzing the impact of feature selection on the accuracy of heart disease prediction | Heart Disease has become one of the most serious diseases that has a
significant impact on human life. It has emerged as one of the leading causes
of mortality among the people across the globe during the last decade. In order
to prevent patients from further damage, an accurate diagnosis of heart disease
on time is an essential factor. Recently we have seen the usage of non-invasive
medical procedures, such as artificial intelligence-based techniques in the
field of medical. Specially machine learning employs several algorithms and
techniques that are widely used and are highly useful in accurately diagnosing
the heart disease with less amount of time. However, the prediction of heart
disease is not an easy task. The increasing size of medical datasets has made
it a complicated task for practitioners to understand the complex feature
relations and make disease predictions. Accordingly, the aim of this research
is to identify the most important risk-factors from a highly dimensional
dataset which helps in the accurate classification of heart disease with less
complications. For a broader analysis, we have used two heart disease datasets
with various medical features. The classification results of the benchmarked
models proved that there is a high impact of relevant features on the
classification accuracy. Even with a reduced number of features, the
performance of the classification models improved significantly with a reduced
training time as compared with models trained on full feature set. |
2,583 | An Analysis of Selection Bias Issue for Online Advertising | In online advertising, a set of potential advertisements can be ranked by a
certain auction system where usually the top-1 advertisement would be selected
and displayed at an advertising space. In this paper, we show a selection bias
issue that is present in an auction system. We analyze that the selection bias
destroy truthfulness of the auction, which implies that the buyers
(advertisers) on the auction can not maximize their profits. Although selection
bias is well known in the field of statistics and there are lot of studies for
it, our main contribution is to combine the theoretical analysis of the bias
with the auction mechanism. In our experiment using online A/B testing, we
evaluate the selection bias on an auction system whose ranking score is the
function of predicted CTR (click through rate) of advertisement. The experiment
showed that the selection bias is drastically reduced by using a multi-task
learning which learns the data for all advertisements. |
2,584 | Inferring Unfairness and Error from Population Statistics in Binary and Multiclass Classification | We propose methods for making inferences on the fairness and accuracy of a
given classifier, using only aggregate population statistics. This is necessary
when it is impossible to obtain individual classification data, for instance
when there is no access to the classifier or to a representative
individual-level validation set. We study fairness with respect to the
equalized odds criterion, which we generalize to multiclass classification. We
propose a measure of unfairness with respect to this criterion, which
quantifies the fraction of the population that is treated unfairly. We then
show how inferences on the unfairness and error of a given classifier can be
obtained using only aggregate label statistics such as the rate of prediction
of each label in each sub-population, as well as the true rate of each label.
We derive inference procedures for binary classifiers and for multiclass
classifiers, for the case where confusion matrices in each sub-population are
known, and for the significantly more challenging case where they are unknown.
We report experiments on data sets representing diverse applications, which
demonstrate the effectiveness and the wide range of possible uses of the
proposed methodology. |
2,585 | Shedding a PAC-Bayesian Light on Adaptive Sliced-Wasserstein Distances | The Sliced-Wasserstein distance (SW) is a computationally efficient and
theoretically grounded alternative to the Wasserstein distance. Yet, the
literature on its statistical properties with respect to the distribution of
slices, beyond the uniform measure, is scarce. To bring new contributions to
this line of research, we leverage the PAC-Bayesian theory and the central
observation that SW actually hinges on a slice-distribution-dependent Gibbs
risk, the kind of quantity PAC-Bayesian bounds have been designed to
characterize. We provide four types of results: i) PAC-Bayesian generalization
bounds that hold on what we refer as adaptive Sliced-Wasserstein distances,
i.e. distances defined with respect to any distribution of slices, ii) a
procedure to learn the distribution of slices that yields a maximally
discriminative SW, by optimizing our PAC-Bayesian bounds, iii) an insight on
how the performance of the so-called distributional Sliced-Wasserstein distance
may be explained through our theory, and iv) empirical illustrations of our
findings. |
2,586 | Improved Cardiac Arrhythmia Prediction Based on Heart Rate Variability Analysis | Many types of ventricular and atrial cardiac arrhythmias have been discovered
in clinical practice in the past 100 years, and these arrhythmias are a major
contributor to sudden cardiac death. Ventricular tachycardia, ventricular
fibrillation, and paroxysmal atrial fibrillation are the most
commonly-occurring and dangerous arrhythmias, therefore early detection is
crucial to prevent any further complications and reduce fatalities. Implantable
devices such as pacemakers are commonly used in patients at high risk of sudden
cardiac death. While great advances have been made in medical technology, there
remain significant challenges in effective management of common arrhythmias.
This thesis proposes novel arrhythmia detection and prediction methods to
differentiate cardiac arrhythmias from non-life-threatening cardiac events, to
increase the likelihood of detecting events that may lead to mortality, as well
as reduce the incidence of unnecessary therapeutic intervention. The methods
are based on detailed analysis of Heart Rate Variability (HRV) information. The
results of the work show good performance of the proposed methods and support
the potential for their deployment in resource-constrained devices for
ventricular and atrial arrhythmia prediction, such as implantable pacemakers
and defibrillators. |
2,587 | Deep Neural Patchworks: Coping with Large Segmentation Tasks | Convolutional neural networks are the way to solve arbitrary image
segmentation tasks. However, when images are large, memory demands often exceed
the available resources, in particular on a common GPU. Especially in
biomedical imaging, where 3D images are common, the problems are apparent. A
typical approach to solve this limitation is to break the task into smaller
subtasks by dividing images into smaller image patches. Another approach, if
applicable, is to look at the 2D image sections separately, and to solve the
problem in 2D. Often, the loss of global context makes such approaches less
effective; important global information might not be present in the current
image patch, or the selected 2D image section. Here, we propose Deep Neural
Patchworks (DNP), a segmentation framework that is based on hierarchical and
nested stacking of patch-based networks that solves the dilemma between global
context and memory limitations. |
2,588 | From "Where" to "What": Towards Human-Understandable Explanations through Concept Relevance Propagation | The emerging field of eXplainable Artificial Intelligence (XAI) aims to bring
transparency to today's powerful but opaque deep learning models. While local
XAI methods explain individual predictions in form of attribution maps, thereby
identifying where important features occur (but not providing information about
what they represent), global explanation techniques visualize what concepts a
model has generally learned to encode. Both types of methods thus only provide
partial insights and leave the burden of interpreting the model's reasoning to
the user. Only few contemporary techniques aim at combining the principles
behind both local and global XAI for obtaining more informative explanations.
Those methods, however, are often limited to specific model architectures or
impose additional requirements on training regimes or data and label
availability, which renders the post-hoc application to arbitrarily pre-trained
models practically impossible. In this work we introduce the Concept Relevance
Propagation (CRP) approach, which combines the local and global perspectives of
XAI and thus allows answering both the "where" and "what" questions for
individual predictions, without additional constraints imposed. We further
introduce the principle of Relevance Maximization for finding representative
examples of encoded concepts based on their usefulness to the model. We thereby
lift the dependency on the common practice of Activation Maximization and its
limitations. We demonstrate the capabilities of our methods in various
settings, showcasing that Concept Relevance Propagation and Relevance
Maximization lead to more human interpretable explanations and provide deep
insights into the model's representations and reasoning through concept
atlases, concept composition analyses, and quantitative investigations of
concept subspaces and their role in fine-grained decision making. |
2,589 | FairVFL: A Fair Vertical Federated Learning Framework with Contrastive Adversarial Learning | Vertical federated learning (VFL) is a privacy-preserving machine learning
paradigm that can learn models from features distributed on different platforms
in a privacy-preserving way. Since in real-world applications the data may
contain bias on fairness-sensitive features (e.g., gender), VFL models may
inherit bias from training data and become unfair for some user groups.
However, existing fair ML methods usually rely on the centralized storage of
fairness-sensitive features to achieve model fairness, which are usually
inapplicable in federated scenarios. In this paper, we propose a fair vertical
federated learning framework (FairVFL), which can improve the fairness of VFL
models. The core idea of FairVFL is to learn unified and fair representations
of samples based on the decentralized feature fields in a privacy-preserving
way. Specifically, each platform with fairness-insensitive features first
learns local data representations from local features. Then, these local
representations are uploaded to a server and aggregated into a unified
representation for the target task. In order to learn fair unified
representations, we send them to each platform storing fairness-sensitive
features and apply adversarial learning to remove bias from the unified
representations inherited from the biased data. Moreover, for protecting user
privacy, we further propose a contrastive adversarial learning method to remove
privacy information from the unified representations in server before sending
them to the platforms keeping fairness-sensitive features. Experiments on two
real-world datasets validate that our method can effectively improve model
fairness with user privacy well-protected. |
2,590 | Explaining the physics of transfer learning a data-driven subgrid-scale closure to a different turbulent flow | Transfer learning (TL) is becoming a powerful tool in scientific applications
of neural networks (NNs), such as weather/climate prediction and turbulence
modeling. TL enables out-of-distribution generalization (e.g., extrapolation in
parameters) and effective blending of disparate training sets (e.g.,
simulations and observations). In TL, selected layers of a NN, already trained
for a base system, are re-trained using a small dataset from a target system.
For effective TL, we need to know 1) what are the best layers to re-train? and
2) what physics are learned during TL? Here, we present novel analyses and a
new framework to address (1)-(2) for a broad range of multi-scale, nonlinear
systems. Our approach combines spectral analyses of the systems' data with
spectral analyses of convolutional NN's activations and kernels, explaining the
inner-workings of TL in terms of the system's nonlinear physics. Using
subgrid-scale modeling of several setups of 2D turbulence as test cases, we
show that the learned kernels are combinations of low-, band-, and high-pass
filters, and that TL learns new filters whose nature is consistent with the
spectral differences of base and target systems. We also find the shallowest
layers are the best to re-train in these cases, which is against the common
wisdom guiding TL in machine learning literature. Our framework identifies the
best layer(s) to re-train beforehand, based on physics and NN theory. Together,
these analyses explain the physics learned in TL and provide a framework to
guide TL for wide-ranging applications in science and engineering, such as
climate change modeling. |
2,591 | Generalized Data Distribution Iteration | To obtain higher sample efficiency and superior final performance
simultaneously has been one of the major challenges for deep reinforcement
learning (DRL). Previous work could handle one of these challenges but
typically failed to address them concurrently. In this paper, we try to tackle
these two challenges simultaneously. To achieve this, we firstly decouple these
challenges into two classic RL problems: data richness and
exploration-exploitation trade-off. Then, we cast these two problems into the
training data distribution optimization problem, namely to obtain desired
training data within limited interactions, and address them concurrently via i)
explicit modeling and control of the capacity and diversity of behavior policy
and ii) more fine-grained and adaptive control of selective/sampling
distribution of the behavior policy using a monotonic data distribution
optimization. Finally, we integrate this process into Generalized Policy
Iteration (GPI) and obtain a more general framework called Generalized Data
Distribution Iteration (GDI). We use the GDI framework to introduce
operator-based versions of well-known RL methods from DQN to Agent57.
Theoretical guarantee of the superiority of GDI compared with GPI is concluded.
We also demonstrate our state-of-the-art (SOTA) performance on Arcade Learning
Environment (ALE), wherein our algorithm has achieved 9620.33% mean human
normalized score (HNS), 1146.39% median HNS and surpassed 22 human world
records using only 200M training frames. Our performance is comparable to
Agent57's while we consume 500 times less data. We argue that there is still a
long way to go before obtaining real superhuman agents in ALE. |
2,592 | Asymptotic Stability in Reservoir Computing | Reservoir Computing is a class of Recurrent Neural Networks with internal
weights fixed at random. Stability relates to the sensitivity of the network
state to perturbations. It is an important property in Reservoir Computing as
it directly impacts performance. In practice, it is desirable to stay in a
stable regime, where the effect of perturbations does not explode
exponentially, but also close to the chaotic frontier where reservoir dynamics
are rich. Open questions remain today regarding input regularization and
discontinuous activation functions. In this work, we use the recurrent kernel
limit to draw new insights on stability in reservoir computing. This limit
corresponds to large reservoir sizes, and it already becomes relevant for
reservoirs with a few hundred neurons. We obtain a quantitative
characterization of the frontier between stability and chaos, which can greatly
benefit hyperparameter tuning. In a broader sense, our results contribute to
understanding the complex dynamics of Recurrent Neural Networks. |
2,593 | A new Hyper-heuristic based on Adaptive Simulated Annealing and Reinforcement Learning for the Capacitated Electric Vehicle Routing Problem | Electric vehicles (EVs) have been adopted in urban areas to reduce
environmental pollution and global warming as a result of the increasing number
of freight vehicles. However, there are still deficiencies in routing the
trajectories of last-mile logistics that continue to impact social and economic
sustainability. For that reason, in this paper, a hyper-heuristic (HH) approach
called Hyper-heuristic Adaptive Simulated Annealing with Reinforcement Learning
(HHASA$_{RL}$) is proposed. It is composed of a multi-armed bandit method and
the self-adaptive Simulated Annealing (SA) metaheuristic algorithm for solving
the problem called Capacitated Electric Vehicle Routing Problem (CEVRP). Due to
the limited number of charging stations and the travel range of EVs, the EVs
must require battery recharging moments in advance and reduce travel times and
costs. The HH implemented improves multiple minimum best-known solutions and
obtains the best mean values for some high-dimensional instances for the
proposed benchmark for the IEEE WCCI2020 competition. |
2,594 | Risk Measures and Upper Probabilities: Coherence and Stratification | Machine learning typically presupposes classical probability theory which
implies that aggregation is built upon expectation. There are now multiple
reasons to motivate looking at richer alternatives to classical probability
theory as a mathematical foundation for machine learning. We systematically
examine a powerful and rich class of such alternatives, known variously as
spectral risk measures, Choquet integrals or Lorentz norms. We present a range
of characterization results, and demonstrate what makes this spectral family so
special. In doing so we demonstrate a natural stratification of all coherent
risk measures in terms of the upper probabilities that they induce by
exploiting results from the theory of rearrangement invariant Banach spaces. We
empirically demonstrate how this new approach to uncertainty helps tackling
practical machine learning problems. |
2,595 | Fooling Explanations in Text Classifiers | State-of-the-art text classification models are becoming increasingly reliant
on deep neural networks (DNNs). Due to their black-box nature, faithful and
robust explanation methods need to accompany classifiers for deployment in
real-life scenarios. However, it has been shown in vision applications that
explanation methods are susceptible to local, imperceptible perturbations that
can significantly alter the explanations without changing the predicted
classes. We show here that the existence of such perturbations extends to text
classifiers as well. Specifically, we introduceTextExplanationFooler (TEF), a
novel explanation attack algorithm that alters text input samples imperceptibly
so that the outcome of widely-used explanation methods changes considerably
while leaving classifier predictions unchanged. We evaluate the performance of
the attribution robustness estimation performance in TEF on five sequence
classification datasets, utilizing three DNN architectures and three
transformer architectures for each dataset. TEF can significantly decrease the
correlation between unchanged and perturbed input attributions, which shows
that all models and explanation methods are susceptible to TEF perturbations.
Moreover, we evaluate how the perturbations transfer to other model
architectures and attribution methods, and show that TEF perturbations are also
effective in scenarios where the target model and explanation method are
unknown. Finally, we introduce a semi-universal attack that is able to compute
fast, computationally light perturbations with no knowledge of the attacked
classifier nor explanation method. Overall, our work shows that explanations in
text classifiers are very fragile and users need to carefully address their
robustness before relying on them in critical applications. |
2,596 | Learning Attention-based Representations from Multiple Patterns for Relation Prediction in Knowledge Graphs | Knowledge bases, and their representations in the form of knowledge graphs
(KGs), are naturally incomplete. Since scientific and industrial applications
have extensively adopted them, there is a high demand for solutions that
complete their information. Several recent works tackle this challenge by
learning embeddings for entities and relations, then employing them to predict
new relations among the entities. Despite their aggrandizement, most of those
methods focus only on the local neighbors of a relation to learn the
embeddings. As a result, they may fail to capture the KGs' context information
by neglecting long-term dependencies and the propagation of entities'
semantics. In this manuscript, we propose {\AE}MP (Attention-based Embeddings
from Multiple Patterns), a novel model for learning contextualized
representations by: (i) acquiring entities' context information through an
attention-enhanced message-passing scheme, which captures the entities' local
semantics while focusing on different aspects of their neighborhood; and (ii)
capturing the semantic context, by leveraging the paths and their relationships
between entities. Our empirical findings draw insights into how attention
mechanisms can improve entities' context representation and how combining
entities and semantic path contexts improves the general representation of
entities and the relation predictions. Experimental results on several large
and small knowledge graph benchmarks show that {\AE}MP either outperforms or
competes with state-of-the-art relation prediction methods. |
2,597 | Look Back When Surprised: Stabilizing Reverse Experience Replay for Neural Approximation | Experience replay methods, which are an essential part of reinforcement
learning(RL) algorithms, are designed to mitigate spurious correlations and
biases while learning from temporally dependent data. Roughly speaking, these
methods allow us to draw batched data from a large buffer such that these
temporal correlations do not hinder the performance of descent algorithms. In
this experimental work, we consider the recently developed and theoretically
rigorous reverse experience replay (RER), which has been shown to remove such
spurious biases in simplified theoretical settings. We combine RER with
optimistic experience replay (OER) to obtain RER++, which is stable under
neural function approximation. We show via experiments that this has a better
performance than techniques like prioritized experience replay (PER) on various
tasks, with a significantly smaller computational complexity. It is well known
in the RL literature that choosing examples greedily with the largest TD error
(as in OER) or forming mini-batches with consecutive data points (as in RER)
leads to poor performance. However, our method, which combines these
techniques, works very well. |
2,598 | Decentralized Low-Latency Collaborative Inference via Ensembles on the Edge | The success of deep neural networks (DNNs) is heavily dependent on
computational resources. While DNNs are often employed on cloud servers, there
is a growing need to operate DNNs on edge devices. Edge devices are typically
limited in their computational resources, yet, often multiple edge devices are
deployed in the same environment and can reliably communicate with each other.
In this work we propose to facilitate the application of DNNs on the edge by
allowing multiple users to collaborate during inference to improve their
accuracy. Our mechanism, coined {\em edge ensembles}, is based on having
diverse predictors at each device, which form an ensemble of models during
inference. To mitigate the communication overhead, the users share quantized
features, and we propose a method for aggregating multiple decisions into a
single inference rule. We analyze the latency induced by edge ensembles,
showing that its performance improvement comes at the cost of a minor
additional delay under common assumptions on the communication network. Our
experiments demonstrate that collaborative inference via edge ensembles
equipped with compact DNNs substantially improves the accuracy over having each
user infer locally, and can outperform using a single centralized DNN larger
than all the networks in the ensemble together. |
2,599 | Utility of Equivariant Message Passing in Cortical Mesh Segmentation | The automated segmentation of cortical areas has been a long-standing
challenge in medical image analysis. The complex geometry of the cortex is
commonly represented as a polygon mesh, whose segmentation can be addressed by
graph-based learning methods. When cortical meshes are misaligned across
subjects, current methods produce significantly worse segmentation results,
limiting their ability to handle multi-domain data. In this paper, we
investigate the utility of E(n)-equivariant graph neural networks (EGNNs),
comparing their performance against plain graph neural networks (GNNs). Our
evaluation shows that GNNs outperform EGNNs on aligned meshes, due to their
ability to leverage the presence of a global coordinate system. On misaligned
meshes, the performance of plain GNNs drop considerably, while E(n)-equivariant
message passing maintains the same segmentation results. The best results can
also be obtained by using plain GNNs on realigned data (co-registered meshes in
a global coordinate system). |
Subsets and Splits
No saved queries yet
Save your SQL queries to embed, download, and access them later. Queries will appear here once saved.