Unnamed: 0
int64 0
5k
| title
stringlengths 9
210
| abstract
stringlengths 164
1.92k
|
|---|---|---|
3,300 | One Loss for Quantization: Deep Hashing with Discrete Wasserstein Distributional Matching | Image hashing is a principled approximate nearest neighbor approach to find
similar items to a query in a large collection of images. Hashing aims to learn
a binary-output function that maps an image to a binary vector. For optimal
retrieval performance, producing balanced hash codes with low-quantization
error to bridge the gap between the learning stage's continuous relaxation and
the inference stage's discrete quantization is important. However, in the
existing deep supervised hashing methods, coding balance and low-quantization
error are difficult to achieve and involve several losses. We argue that this
is because the existing quantization approaches in these methods are
heuristically constructed and not effective to achieve these objectives. This
paper considers an alternative approach to learning the quantization
constraints. The task of learning balanced codes with low quantization error is
re-formulated as matching the learned distribution of the continuous codes to a
pre-defined discrete, uniform distribution. This is equivalent to minimizing
the distance between two distributions. We then propose a computationally
efficient distributional distance by leveraging the discrete property of the
hash functions. This distributional distance is a valid distance and enjoys
lower time and sample complexities. The proposed single-loss quantization
objective can be integrated into any existing supervised hashing method to
improve code balance and quantization error. Experiments confirm that the
proposed approach substantially improves the performance of several
representative hashing~methods. |
3,301 | Multi-Agent Learning of Numerical Methods for Hyperbolic PDEs with Factored Dec-MDP | Factored decentralized Markov decision process (Dec-MDP) is a framework for
modeling sequential decision making problems in multi-agent systems. In this
paper, we formalize the learning of numerical methods for hyperbolic partial
differential equations (PDEs), specifically the Weighted Essentially
Non-Oscillatory (WENO) scheme, as a factored Dec-MDP problem. We show that
different reward formulations lead to either reinforcement learning (RL) or
behavior cloning, and a homogeneous policy could be learned for all agents
under the RL formulation with a policy gradient algorithm. Because the trained
agents only act on their local observations, the multi-agent system can be used
as a general numerical method for hyperbolic PDEs and generalize to different
spatial discretizations, episode lengths, dimensions, and even equation types. |
3,302 | Multilingual Transformers for Product Matching -- Experiments and a New Benchmark in Polish | Product matching corresponds to the task of matching identical products
across different data sources. It typically employs available product features
which, apart from being multimodal, i.e., comprised of various data types,
might be non-homogeneous and incomplete. The paper shows that pre-trained,
multilingual Transformer models, after fine-tuning, are suitable for solving
the product matching problem using textual features both in English and Polish
languages. We tested multilingual mBERT and XLM-RoBERTa models in English on
Web Data Commons - training dataset and gold standard for large-scale product
matching. The obtained results show that these models perform similarly to the
latest solutions tested on this set, and in some cases, the results were even
better.
Additionally, we prepared a new dataset entirely in Polish and based on
offers in selected categories obtained from several online stores for the
research purpose. It is the first open dataset for product matching tasks in
Polish, which allows comparing the effectiveness of the pre-trained models.
Thus, we also showed the baseline results obtained by the fine-tuned mBERT and
XLM-RoBERTa models on the Polish datasets. |
3,303 | Mitigating Dataset Bias by Using Per-sample Gradient | The performance of deep neural networks is strongly influenced by the
training dataset setup. In particular, when attributes having a strong
correlation with the target attribute are present, the trained model can
provide unintended prejudgments and show significant inference errors (i.e.,
the dataset bias problem). Various methods have been proposed to mitigate
dataset bias, and their emphasis is on weakly correlated samples, called
bias-conflicting samples. These methods are based on explicit bias labels
involving human or empirical correlation metrics (e.g., training loss).
However, such metrics require human costs or have insufficient theoretical
explanation. In this study, we propose a debiasing algorithm, called PGD
(Per-sample Gradient-based Debiasing), that comprises three steps: (1) training
a model on uniform batch sampling, (2) setting the importance of each sample in
proportion to the norm of the sample gradient, and (3) training the model using
importance-batch sampling, whose probability is obtained in step (2). Compared
with existing baselines for various synthetic and real-world datasets, the
proposed method showed state-of-the-art accuracy for a the classification task.
Furthermore, we describe theoretical understandings about how PGD can mitigate
dataset bias. |
3,304 | Lessons Learned from Data-Driven Building Control Experiments: Contrasting Gaussian Process-based MPC, Bilevel DeePC, and Deep Reinforcement Learning | This manuscript offers the perspective of experimentalists on a number of
modern data-driven techniques: model predictive control relying on Gaussian
processes, adaptive data-driven control based on behavioral theory, and deep
reinforcement learning. These techniques are compared in terms of data
requirements, ease of use, computational burden, and robustness in the context
of real-world applications. Our remarks and observations stem from a number of
experimental investigations carried out in the field of building control in
diverse environments, from lecture halls and apartment spaces to a hospital
surgery center. The final goal is to support others in identifying what
technique is best suited to tackle their own problems. |
3,305 | Provable General Function Class Representation Learning in Multitask Bandits and MDPs | While multitask representation learning has become a popular approach in
reinforcement learning (RL) to boost the sample efficiency, the theoretical
understanding of why and how it works is still limited. Most previous
analytical works could only assume that the representation function is already
known to the agent or from linear function class, since analyzing general
function class representation encounters non-trivial technical obstacles such
as generalization guarantee, formulation of confidence bound in abstract
function space, etc. However, linear-case analysis heavily relies on the
particularity of linear function class, while real-world practice usually
adopts general non-linear representation functions like neural networks. This
significantly reduces its applicability. In this work, we extend the analysis
to general function class representations. Specifically, we consider an agent
playing $M$ contextual bandits (or MDPs) concurrently and extracting a shared
representation function $\phi$ from a specific function class $\Phi$ using our
proposed Generalized Functional Upper Confidence Bound algorithm (GFUCB). We
theoretically validate the benefit of multitask representation learning within
general function class for bandits and linear MDP for the first time. Lastly,
we conduct experiments to demonstrate the effectiveness of our algorithm with
neural net representation. |
3,306 | A novel approach to rating transition modelling via Machine Learning and SDEs on Lie groups | In this paper, we introduce a novel methodology to model rating transitions
with a stochastic process. To introduce stochastic processes, whose values are
valid rating matrices, we noticed the geometric properties of stochastic
matrices and its link to matrix Lie groups. We give a gentle introduction to
this topic and demonstrate how It\^o-SDEs in R will generate the desired model
for rating transitions. To calibrate the rating model to historical data, we
use a Deep-Neural-Network (DNN) called TimeGAN to learn the features of a time
series of historical rating matrices. Then, we use this DNN to generate
synthetic rating transition matrices. Afterwards, we fit the moments of the
generated rating matrices and the rating process at specific time points, which
results in a good fit. After calibration, we discuss the quality of the
calibrated rating transition process by examining some properties that a time
series of rating matrices should satisfy, and we will see that this geometric
approach works very well. |
3,307 | Static Scheduling with Predictions Learned through Efficient Exploration | A popular approach to go beyond the worst-case analysis of online algorithms
is to assume the existence of predictions that can be leveraged to improve
performances. Those predictions are usually given by some external sources that
cannot be fully trusted. Instead, we argue that trustful predictions can be
built by algorithms, while they run. We investigate this idea in the
illustrative context of static scheduling with exponential job sizes. Indeed,
we prove that algorithms agnostic to this structure do not perform better than
in the worst case. In contrast, when the expected job sizes are known, we show
that the best algorithm using this information, called
Follow-The-Perfect-Prediction (FTPP), exhibits much better performances. Then,
we introduce two adaptive explore-then-commit types of algorithms: they both
first (partially) learn expected job sizes and then follow FTPP once their
self-predictions are confident enough. On the one hand, ETCU explores in
"series", by completing jobs sequentially to acquire information. On the other
hand, ETCRR, inspired by the optimal worst-case algorithm Round-Robin (RR),
explores efficiently in "parallel". We prove that both of them asymptotically
reach the performances of FTPP, with a faster rate for ETCRR. Those findings
are empirically evaluated on synthetic data. |
3,308 | Differentiable programming for functional connectomics | Mapping the functional connectome has the potential to uncover key insights
into brain organisation. However, existing workflows for functional
connectomics are limited in their adaptability to new data, and principled
workflow design is a challenging combinatorial problem. We introduce a new
analytic paradigm and software toolbox that implements common operations used
in functional connectomics as fully differentiable processing blocks. Under
this paradigm, workflow configurations exist as reparameterisations of a
differentiable functional that interpolates them. The differentiable program
that we envision occupies a niche midway between traditional pipelines and
end-to-end neural networks, combining the glass-box tractability and domain
knowledge of the former with the amenability to optimisation of the latter. In
this preliminary work, we provide a proof of concept for differentiable
connectomics, demonstrating the capacity of our processing blocks both to
recapitulate canonical knowledge in neuroscience and to make new discoveries in
an unsupervised setting. Our differentiable modules are competitive with
state-of-the-art methods in problem domains including functional parcellation,
denoising, and covariance modelling. Taken together, our results and software
demonstrate the promise of differentiable programming for functional
connectomics. |
3,309 | Simulation-Based Inference with WALDO: Perfectly Calibrated Confidence Regions Using Any Prediction or Posterior Estimation Algorithm | The vast majority of modern machine learning targets prediction problems,
with algorithms such as Deep Neural Networks revolutionizing the accuracy of
point predictions for high-dimensional complex data. Predictive approaches are
now used in many domain sciences to directly estimate internal parameters of
interest in theoretical simulator-based models. In parallel, common
alternatives focus on estimating the full posterior using modern neural density
estimators such as normalizing flows. However, an open problem in
simulation-based inference (SBI) is how to construct properly calibrated
confidence regions for internal parameters with nominal conditional coverage
and high power. Many SBI methods are indeed known to produce overly confident
posterior approximations, yielding misleading uncertainty estimates. Similarly,
existing approaches for uncertainty quantification in deep learning provide no
guarantees on conditional coverage. In this work, we present WALDO, a novel
method for constructing correctly calibrated confidence regions in SBI. WALDO
reframes the well-known Wald test and uses Neyman inversion to convert point
predictions and posteriors from any prediction or posterior estimation
algorithm to confidence sets with correct conditional coverage, even for finite
sample sizes. As a concrete example, we demonstrate how a recently proposed
deep learning prediction approach for particle energies in high-energy physics
can be recalibrated using WALDO to produce confidence intervals with correct
coverage and high power. |
3,310 | Automatic Relation-aware Graph Network Proliferation | Graph neural architecture search has sparked much attention as Graph Neural
Networks (GNNs) have shown powerful reasoning capability in many relational
tasks. However, the currently used graph search space overemphasizes learning
node features and neglects mining hierarchical relational information.
Moreover, due to diverse mechanisms in the message passing, the graph search
space is much larger than that of CNNs. This hinders the straightforward
application of classical search strategies for exploring complicated graph
search space. We propose Automatic Relation-aware Graph Network Proliferation
(ARGNP) for efficiently searching GNNs with a relation-guided message passing
mechanism. Specifically, we first devise a novel dual relation-aware graph
search space that comprises both node and relation learning operations. These
operations can extract hierarchical node/relational information and provide
anisotropic guidance for message passing on a graph. Second, analogous to cell
proliferation, we design a network proliferation search paradigm to
progressively determine the GNN architectures by iteratively performing network
division and differentiation. The experiments on six datasets for four graph
learning tasks demonstrate that GNNs produced by our method are superior to the
current state-of-the-art hand-crafted and search-based GNNs. Codes are
available at https://github.com/phython96/ARGNP. |
3,311 | Augmentation-Aware Self-Supervision for Data-Efficient GAN Training | Training generative adversarial networks (GANs) with limited data is valuable
but challenging because discriminators are prone to over-fitting in such
situations. Recently proposed differentiable data augmentation techniques for
discriminators demonstrate improved data efficiency of training GANs. However,
the naive data augmentation introduces undesired invariance to augmentation
into the discriminator. The invariance may degrade the representation learning
ability of the discriminator, thereby affecting the generative modeling
performance of the generator. To mitigate the invariance while inheriting the
benefits of data augmentation, we propose a novel augmentation-aware
self-supervised discriminator that predicts the parameter of augmentation given
the augmented and original data. Moreover, the prediction task is required to
distinguishable between real data and generated data since they are different
during training. We further encourage the generator to learn from the proposed
discriminator by generating augmentation-predictable real data. We compare the
proposed method with state-of-the-arts across the class-conditional BigGAN and
unconditional StyleGAN2 architectures on CIFAR-10/100 and several low-shot
datasets, respectively. Experimental results show a significantly improved
generation performance of our method over competing methods for training
data-efficient GANs. |
3,312 | Contrastive Representation Learning for 3D Protein Structures | Learning from 3D protein structures has gained wide interest in protein
modeling and structural bioinformatics. Unfortunately, the number of available
structures is orders of magnitude lower than the training data sizes commonly
used in computer vision and machine learning. Moreover, this number is reduced
even further, when only annotated protein structures can be considered, making
the training of existing models difficult and prone to over-fitting. To address
this challenge, we introduce a new representation learning framework for 3D
protein structures. Our framework uses unsupervised contrastive learning to
learn meaningful representations of protein structures, making use of proteins
from the Protein Data Bank. We show, how these representations can be used to
solve a large variety of tasks, such as protein function prediction, protein
fold classification, structural similarity prediction, and protein-ligand
binding affinity prediction. Moreover, we show how fine-tuned networks,
pre-trained with our algorithm, lead to significantly improved task
performance, achieving new state-of-the-art results in many tasks. |
3,313 | Generalised Implicit Neural Representations | We consider the problem of learning implicit neural representations (INRs)
for signals on non-Euclidean domains. In the Euclidean case, INRs are trained
on a discrete sampling of a signal over a regular lattice. Here, we assume that
the continuous signal exists on some unknown topological space from which we
sample a discrete graph. In the absence of a coordinate system to identify the
sampled nodes, we propose approximating their location with a spectral
embedding of the graph. This allows us to train INRs without knowing the
underlying continuous domain, which is the case for most graph signals in
nature, while also making the INRs equivariant under the symmetry group of the
domain. We show experiments with our method on various real-world signals on
non-Euclidean domains. |
3,314 | COIN: Co-Cluster Infomax for Bipartite Graphs | Bipartite graphs are powerful data structures to model interactions between
two types of nodes, which have been used in a variety of applications, such as
recommender systems, information retrieval, and drug discovery. A fundamental
challenge for bipartite graphs is how to learn informative node embeddings.
Despite the success of recent self-supervised learning methods on bipartite
graphs, their objectives are discriminating instance-wise positive and negative
node pairs, which could contain cluster-level errors. In this paper, we
introduce a novel co-cluster infomax (COIN) framework, which captures the
cluster-level information by maximizing the mutual information of co-clusters.
Different from previous infomax methods which estimate mutual information by
neural networks, COIN could easily calculate mutual information. Besides, COIN
is an end-to-end co-clustering method which can be trained jointly with other
objective functions and optimized via back-propagation. Furthermore, we also
provide theoretical analysis for COIN. We theoretically prove that COIN is able
to effectively maximize the mutual information of node embeddings and COIN is
upper-bounded by the prior distributions of nodes. We extensively evaluate the
proposed COIN framework on various benchmark datasets and tasks to demonstrate
the effectiveness of COIN. |
3,315 | Multi-task Optimization Based Co-training for Electricity Consumption Prediction | Real-world electricity consumption prediction may involve different tasks,
e.g., prediction for different time steps ahead or different geo-locations.
These tasks are often solved independently without utilizing some common
problem-solving knowledge that could be extracted and shared among these tasks
to augment the performance of solving each task. In this work, we propose a
multi-task optimization (MTO) based co-training (MTO-CT) framework, where the
models for solving different tasks are co-trained via an MTO paradigm in which
solving each task may benefit from the knowledge gained from when solving some
other tasks to help its solving process. MTO-CT leverages long short-term
memory (LSTM) based model as the predictor where the knowledge is represented
via connection weights and biases. In MTO-CT, an inter-task knowledge transfer
module is designed to transfer knowledge between different tasks, where the
most helpful source tasks are selected by using the probability matching and
stochastic universal selection, and evolutionary operations like mutation and
crossover are performed for reusing the knowledge from selected source tasks in
a target task. We use electricity consumption data from five states in
Australia to design two sets of tasks at different scales: a) one-step ahead
prediction for each state (five tasks) and b) 6-step, 12-step, 18-step, and
24-step ahead prediction for each state (20 tasks). The performance of MTO-CT
is evaluated on solving each of these two sets of tasks in comparison to
solving each task in the set independently without knowledge sharing under the
same settings, which demonstrates the superiority of MTO-CT in terms of
prediction accuracy. |
3,316 | The CLRS Algorithmic Reasoning Benchmark | Learning representations of algorithms is an emerging area of machine
learning, seeking to bridge concepts from neural networks with classical
algorithms. Several important works have investigated whether neural networks
can effectively reason like algorithms, typically by learning to execute them.
The common trend in the area, however, is to generate targeted kinds of
algorithmic data to evaluate specific hypotheses, making results hard to
transfer across publications, and increasing the barrier of entry. To
consolidate progress and work towards unified evaluation, we propose the CLRS
Algorithmic Reasoning Benchmark, covering classical algorithms from the
Introduction to Algorithms textbook. Our benchmark spans a variety of
algorithmic reasoning procedures, including sorting, searching, dynamic
programming, graph algorithms, string algorithms and geometric algorithms. We
perform extensive experiments to demonstrate how several popular algorithmic
reasoning baselines perform on these tasks, and consequently, highlight links
to several open challenges. Our library is readily available at
https://github.com/deepmind/clrs. |
3,317 | Sample-Efficient, Exploration-Based Policy Optimisation for Routing Problems | Model-free deep-reinforcement-based learning algorithms have been applied to
a range of
COPs~\cite{bello2016neural}~\cite{kool2018attention}~\cite{nazari2018reinforcement}.
However, these approaches suffer from two key challenges when applied to
combinatorial problems: insufficient exploration and the requirement of many
training examples of the search space to achieve reasonable performance.
Combinatorial optimisation can be complex, characterised by search spaces with
many optimas and large spaces to search and learn. Therefore, a new method is
needed to find good solutions that are more efficient by being more sample
efficient. This paper presents a new reinforcement learning approach that is
based on entropy. In addition, we design an off-policy-based reinforcement
learning technique that maximises the expected return and improves the sample
efficiency to achieve faster learning during training time. We systematically
evaluate our approach on a range of route optimisation tasks typically used to
evaluate learning-based optimisation, such as the such as the Travelling
Salesman problems (TSP), Capacitated Vehicle Routing Problem (CVRP). In this
paper, we show that our model can generalise to various route problems, such as
the split-delivery VRP (SDVRP), and compare the performance of our method with
that of current state-of-the-art approaches. The Empirical results show that
the proposed method can improve on state-of-the-art methods in terms of
solution quality and computation time and generalise to problems of different
sizes. |
3,318 | Label-Enhanced Graph Neural Network for Semi-supervised Node Classification | Graph Neural Networks (GNNs) have been widely applied in the semi-supervised
node classification task, where a key point lies in how to sufficiently
leverage the limited but valuable label information. Most of the classical GNNs
solely use the known labels for computing the classification loss at the
output. In recent years, several methods have been designed to additionally
utilize the labels at the input. One part of the methods augment the node
features via concatenating or adding them with the one-hot encodings of labels,
while other methods optimize the graph structure by assuming neighboring nodes
tend to have the same label. To bring into full play the rich information of
labels, in this paper, we present a label-enhanced learning framework for GNNs,
which first models each label as a virtual center for intra-class nodes and
then jointly learns the representations of both nodes and labels. Our approach
could not only smooth the representations of nodes belonging to the same class,
but also explicitly encode the label semantics into the learning process of
GNNs. Moreover, a training node selection technique is provided to eliminate
the potential label leakage issue and guarantee the model generalization
ability. Finally, an adaptive self-training strategy is proposed to iteratively
enlarge the training set with more reliable pseudo labels and distinguish the
importance of each pseudo-labeled node during the model training process.
Experimental results on both real-world and synthetic datasets demonstrate our
approach can not only consistently outperform the state-of-the-arts, but also
effectively smooth the representations of intra-class nodes. |
3,319 | Are classical neural networks quantum? | Neural networks are being used to improve the probing of the state spaces of
many particle systems as approximations to wavefunctions and in order to avoid
the recurring sign problem of quantum monte-carlo. One may ask whether the
usual classical neural networks have some actual hidden quantum properties that
make them such suitable tools for a highly coupled quantum problem. I discuss
here what makes a system quantum and to what extent we can interpret a neural
network as having quantum remnants. |
3,320 | Differentiable Invariant Causal Discovery | Learning causal structure from observational data is a fundamental challenge
in machine learning. The majority of commonly used differentiable causal
discovery methods are non-identifiable, turning this problem into a continuous
optimization task prone to data biases. In many real-life situations, data is
collected from different environments, in which the functional relations remain
consistent across environments, while the distribution of additive noises may
vary. This paper proposes Differentiable Invariant Causal Discovery (DICD),
utilizing the multi-environment information based on a differentiable framework
to avoid learning spurious edges and wrong causal directions. Specifically,
DICD aims to discover the environment-invariant causation while removing the
environment-dependent correlation. We further formulate the constraint that
enforces the target structure equation model to maintain optimal across the
environments. Theoretical guarantees for the identifiability of proposed DICD
are provided under mild conditions with enough environments. Extensive
experiments on synthetic and real-world datasets verify that DICD outperforms
state-of-the-art causal discovery methods up to 36% in SHD. Our code will be
open-sourced upon acceptance. |
3,321 | Improvements to Supervised EM Learning of Shared Kernel Models by Feature Space Partitioning | Expectation maximisation (EM) is usually thought of as an unsupervised
learning method for estimating the parameters of a mixture distribution,
however it can also be used for supervised learning when class labels are
available. As such, EM has been applied to train neural nets including the
probabilistic radial basis function (PRBF) network or shared kernel (SK) model.
This paper addresses two major shortcomings of previous work in this area: the
lack of rigour in the derivation of the EM training algorithm; and the
computational complexity of the technique, which has limited it to low
dimensional data sets. We first present a detailed derivation of EM for the
Gaussian shared kernel model PRBF classifier, making use of data association
theory to obtain the complete data likelihood, Baum's auxiliary function (the
E-step) and its subsequent maximisation (M-step). To reduce complexity of the
resulting SKEM algorithm, we partition the feature space into $R$
non-overlapping subsets of variables. The resulting product decomposition of
the joint data likelihood, which is exact when the feature partitions are
independent, allows the SKEM to be implemented in parallel and at $R^2$ times
lower complexity. The operation of the partitioned SKEM algorithm is
demonstrated on the MNIST data set and compared with its non-partitioned
counterpart. It eventuates that improved performance at reduced complexity is
achievable. Comparisons with standard classification algorithms are provided on
a number of other benchmark data sets. |
3,322 | Scalable Distributional Robustness in a Class of Non Convex Optimization with Guarantees | Distributionally robust optimization (DRO) has shown lot of promise in
providing robustness in learning as well as sample based optimization problems.
We endeavor to provide DRO solutions for a class of sum of fractionals,
non-convex optimization which is used for decision making in prominent areas
such as facility location and security games. In contrast to previous work, we
find it more tractable to optimize the equivalent variance regularized form of
DRO rather than the minimax form. We transform the variance regularized form to
a mixed-integer second order cone program (MISOCP), which, while guaranteeing
near global optimality, does not scale enough to solve problems with real world
data-sets. We further propose two abstraction approaches based on clustering
and stratified sampling to increase scalability, which we then use for real
world data-sets. Importantly, we provide near global optimality guarantees for
our approach and show experimentally that our solution quality is better than
the locally optimal ones achieved by state-of-the-art gradient-based methods.
We experimentally compare our different approaches and baselines, and reveal
nuanced properties of a DRO solution. |
3,323 | k-Means Maximum Entropy Exploration | Exploration in high-dimensional, continuous spaces with sparse rewards is an
open problem in reinforcement learning. Artificial curiosity algorithms address
this by creating rewards that lead to exploration. Given a reinforcement
learning algorithm capable of maximizing rewards, the problem reduces to
finding an optimization objective consistent with exploration. Maximum entropy
exploration uses the entropy of the state visitation distribution as such an
objective. However, efficiently estimating the entropy of the state visitation
distribution is challenging in high-dimensional, continuous spaces. We
introduce an artificial curiosity algorithm based on lower bounding an
approximation to the entropy of the state visitation distribution. The bound
relies on a result we prove for non-parametric density estimation in arbitrary
dimensions using k-means. We show that our approach is both computationally
efficient and competitive on benchmarks for exploration in high-dimensional,
continuous spaces, especially on tasks where reinforcement learning algorithms
are unable to find rewards. |
3,324 | Meta-ticket: Finding optimal subnetworks for few-shot learning within randomly initialized neural networks | Few-shot learning for neural networks (NNs) is an important problem that aims
to train NNs with a few data. The main challenge is how to avoid overfitting
since over-parameterized NNs can easily overfit to such small dataset. Previous
work (e.g. MAML by Finn et al. 2017) tackles this challenge by meta-learning,
which learns how to learn from a few data by using various tasks. On the other
hand, one conventional approach to avoid overfitting is restricting hypothesis
spaces by endowing sparse NN structures like convolution layers in computer
vision. However, although such manually-designed sparse structures are
sample-efficient for sufficiently large datasets, they are still insufficient
for few-shot learning. Then the following questions naturally arise: (1) Can we
find sparse structures effective for few-shot learning by meta-learning? (2)
What benefits will it bring in terms of meta-generalization? In this work, we
propose a novel meta-learning approach, called Meta-ticket, to find optimal
sparse subnetworks for few-shot learning within randomly initialized NNs. We
empirically validated that Meta-ticket successfully discover sparse subnetworks
that can learn specialized features for each given task. Due to this task-wise
adaptation ability, Meta-ticket achieves superior meta-generalization compared
to MAML-based methods especially with large NNs. |
3,325 | Optimizing Intermediate Representations of Generative Models for Phase Retrieval | Phase retrieval is the problem of reconstructing images from magnitude-only
measurements. In many real-world applications the problem is underdetermined.
When training data is available, generative models are a new idea to constrain
the solution set. However, not all possible solutions are within the range of
the generator. Instead, they are represented with some error. To reduce this
representation error in the context of phase retrieval, we first leverage a
novel variation of intermediate layer optimization (ILO) to extend the range of
the generator while still producing images consistent with the training data.
Second, we introduce new initialization schemes that further improve the
quality of the reconstruction. With extensive experiments on Fourier and
Gaussian phase retrieval problems and thorough ablation studies, we can show
the benefits of our modified ILO and the new initialization schemes. |
3,326 | Communication-Efficient Distributionally Robust Decentralized Learning | Decentralized learning algorithms empower interconnected edge devices to
share data and computational resources to collaboratively train a machine
learning model without the aid of a central coordinator (e.g. an orchestrating
basestation). In the case of heterogeneous data distributions at the network
devices, collaboration can yield predictors with unsatisfactory performance for
a subset of the devices. For this reason, in this work we consider the
formulation of a distributionally robust decentralized learning task and we
propose a decentralized single loop gradient descent/ascent algorithm (AD-GDA)
to solve the underlying minimax optimization problem. We render our algorithm
communication efficient by employing a compressed consensus scheme and we
provide convergence guarantees for smooth convex and non-convex loss functions.
Finally, we corroborate the theoretical findings with empirical evidence of the
ability of the proposed algorithm in providing unbiased predictors over a
network of collaborating devices with highly heterogeneous data distributions. |
3,327 | GlanceNets: Interpretabile, Leak-proof Concept-based Models | There is growing interest in concept-based models (CBMs) that combine
high-performance and interpretability by acquiring and reasoning with a
vocabulary of high-level concepts. A key requirement is that the concepts be
interpretable. Existing CBMs tackle this desideratum using a variety of
heuristics based on unclear notions of interpretability, and fail to acquire
concepts with the intended semantics. We address this by providing a clear
definition of interpretability in terms of alignment between the model's
representation and an underlying data generation process, and introduce
GlanceNets, a new CBM that exploits techniques from disentangled representation
learning and open-set recognition to achieve alignment, thus improving the
interpretability of the learned concepts. We show that GlanceNets, paired with
concept-level supervision, achieve better alignment than state-of-the-art
approaches while preventing spurious information from unintendedly leaking into
the learned concepts. |
3,328 | Prompt Injection: Parameterization of Fixed Inputs | Recent works have shown that attaching prompts to the input is effective at
conditioning Language Models (LM) to perform specific tasks. However, prompts
are always included in the input text during inference, thus incurring
substantial computational and memory overhead. Also, there is currently no
straightforward method of utilizing prompts that are longer than the maximum
input length of the LMs without incurring additional costs during inference. We
propose Prompt Injection (PI), a novel formulation of injecting the prompt into
the parameters of an LM to be an efficient alternative to attaching fixed
prompts to the input. We show that in scenarios with long fixed prompts, PI can
be up to 280 times more efficient in terms of total FLOPs than previous
approaches. We further explore methodologies for PI and show promising results
in persona-dependent conversation, semantic parsing, and zero-shot learning
with task instructions. Through these explorations, we show that PI can be a
promising direction for conditioning language models, especially in scenarios
with long and fixed prompts. |
3,329 | Individual health-disease phase diagrams for disease prevention based on machine learning | Early disease detection and prevention methods based on effective
interventions are gaining attention. Machine learning technology has enabled
precise disease prediction by capturing individual differences in multivariate
data. Progress in precision medicine has revealed that substantial
heterogeneity exists in health data at the individual level and that complex
health factors are involved in the development of chronic diseases. However, it
remains a challenge to identify individual physiological state changes in
cross-disease onset processes because of the complex relationships among
multiple biomarkers. Here, we present the health-disease phase diagram (HDPD),
which represents a personal health state by visualizing the boundary values of
multiple biomarkers that fluctuate early in the disease progression process. In
HDPDs, future onset predictions are represented by perturbing multiple
biomarker values while accounting for dependencies among variables. We
constructed HDPDs for 11 non-communicable diseases (NCDs) from a longitudinal
health checkup cohort of 3,238 individuals, comprising 3,215 measurement items
and genetic data. Improvement of biomarker values to the non-onset region in
HDPD significantly prevented future disease onset in 7 out of 11 NCDs. Our
results demonstrate that HDPDs can represent individual physiological states in
the onset process and be used as intervention goals for disease prevention. |
3,330 | Semantic Autoencoder and Its Potential Usage for Adversarial Attack | Autoencoder can give rise to an appropriate latent representation of the
input data, however, the representation which is solely based on the intrinsic
property of the input data, is usually inferior to express some semantic
information. A typical case is the potential incapability of forming a clear
boundary upon clustering of these representations. By encoding the latent
representation that not only depends on the content of the input data, but also
the semantic of the input data, such as label information, we propose an
enhanced autoencoder architecture named semantic autoencoder. Experiments of
representation distribution via t-SNE shows a clear distinction between these
two types of encoders and confirm the supremacy of the semantic one, whilst the
decoded samples of these two types of autoencoders exhibit faint dissimilarity
either objectively or subjectively. Based on this observation, we consider
adversarial attacks to learning algorithms that rely on the latent
representation obtained via autoencoders. It turns out that latent contents of
adversarial samples constructed from semantic encoder with deliberate wrong
label information exhibit different distribution compared with that of the
original input data, while both of these samples manifest very marginal
difference. This new way of attack set up by our work is worthy of attention
due to the necessity to secure the widespread deep learning applications. |
3,331 | Improving Ads-Profitability Using Traffic-Fingerprints | This paper introduces the concept of traffic-fingerprints, i.e., normalized
24-dimensional vectors representing a distribution of daily traffic on a web
page. Using k-means clustering we show that similarity of traffic-fingerprints
is related to the similarity of profitability time patterns for ads shown on
these pages. In other words, these fingerprints are correlated with the
conversions rates, thus allowing us to argue about conversion rates on pages
with negligible traffic. By blocking or unblocking whole clusters of pages we
were able to increase the revenue of online campaigns by more than 50%. |
3,332 | Automatic differentiation of nonsmooth iterative algorithms | Differentiation along algorithms, i.e., piggyback propagation of derivatives,
is now routinely used to differentiate iterative solvers in differentiable
programming. Asymptotics is well understood for many smooth problems but the
nondifferentiable case is hardly considered. Is there a limiting object for
nonsmooth piggyback automatic differentiation (AD)? Does it have any
variational meaning and can it be used effectively in machine learning? Is
there a connection with classical derivative? All these questions are addressed
under appropriate nonexpansivity conditions in the framework of conservative
derivatives which has proved useful in understanding nonsmooth AD. For
nonsmooth piggyback iterations, we characterize the attractor set of nonsmooth
piggyback iterations as a set-valued fixed point which remains in the
conservative framework. This has various consequences and in particular almost
everywhere convergence of classical derivatives. Our results are illustrated on
parametric convex optimization problems with forward-backward, Douglas-Rachford
and Alternating Direction of Multiplier algorithms as well as the Heavy-Ball
method. |
3,333 | Comparing interpretation methods in mental state decoding analyses with deep learning models | Deep learning (DL) methods find increasing application in mental state
decoding, where researchers seek to understand the mapping between mental
states (such as accepting or rejecting a gamble) and brain activity, by
identifying those brain regions (and networks) whose activity allows to
accurately identify (i.e., decode) these states. Once DL models have been
trained to accurately decode a set of mental states, neuroimaging researchers
often make use of interpretation methods from explainable artificial
intelligence research to understand their learned mappings between mental
states and brain activity. Here, we compare the explanations of prominent
interpretation methods for the mental state decoding decisions of DL models
trained on three functional Magnetic Resonance Imaging (fMRI) datasets. We find
that interpretation methods that capture the model's decision process well, by
producing faithful explanations, generally produce explanations that are less
in line with the results of standard analyses of the fMRI data, when compared
to the explanations of interpretation methods with less explanation
faithfulness. Specifically, we find that interpretation methods that focus on
how sensitively a model's decoding decision changes with the values of the
input produce explanations that better match with the results of a standard
general linear model analysis of the fMRI data, while interpretation methods
that focus on identifying the specific contribution of an input feature's value
to the decoding decision produce overall more faithful explanations that align
less well with the results of standard analyses of the fMRI data. |
3,334 | A Computation and Communication Efficient Method for Distributed Nonconvex Problems in the Partial Participation Setting | We present a new method that includes three key components of distributed
optimization and federated learning: variance reduction of stochastic
gradients, compressed communication, and partial participation. We prove that
the new method has optimal oracle complexity and state-of-the-art communication
complexity in the partial participation setting. Moreover, we observe that "1 +
1 + 1 is not 3": by mixing variance reduction of stochastic gradients with
compressed communication and partial participation, we do not obtain a fully
synergetic effect. We explain the nature of this phenomenon, argue that this is
to be expected, and propose possible workarounds. |
3,335 | GSR: A Generalized Symbolic Regression Approach | Identifying the mathematical relationships that best describe a dataset
remains a very challenging problem in machine learning, and is known as
Symbolic Regression (SR). In contrast to neural networks which are often
treated as black boxes, SR attempts to gain insight into the underlying
relationships between the independent variables and the target variable of a
given dataset by assembling analytical functions. In this paper, we present
GSR, a Generalized Symbolic Regression approach, by modifying the conventional
SR optimization problem formulation, while keeping the main SR objective
intact. In GSR, we infer mathematical relationships between the independent
variables and some transformation of the target variable. We constrain our
search space to a weighted sum of basis functions, and propose a genetic
programming approach with a matrix-based encoding scheme. We show that our GSR
method outperforms several state-of-the-art methods on the well-known SR
benchmark problem sets. Finally, we highlight the strengths of GSR by
introducing SymSet, a new SR benchmark set which is more challenging relative
to the existing benchmarks. |
3,336 | HW-Aware Initialization of DNN Auto-Tuning to Improve Exploration Time and Robustness | The process of optimizing the latency of DNN operators with ML models and
hardware-in-the-loop, called auto-tuning, has established itself as a pervasive
method for the deployment of neural networks. From a search space of
loop-optimizations, the candidate providing the best performance has to be
selected. Performance of individual configurations is evaluated through
hardware measurements. The combinatorial explosion of possible configurations,
together with the cost of hardware evaluation makes exhaustive explorations of
the search space infeasible in practice. Machine Learning methods, like random
forests or reinforcement learning are used to aid in the selection of
candidates for hardware evaluation. For general purpose hardware like x86 and
GPGPU architectures impressive performance gains can be achieved, compared to
hand-optimized libraries like cuDNN. The method is also useful in the space of
hardware accelerators with less wide-spread adoption, where a high-performance
library is not always available. However, hardware accelerators are often less
flexible with respect to their programming which leads to operator
configurations not executable on the hardware target. This work evaluates how
these invalid configurations affect the auto-tuning process and its underlying
performance prediction model for the VTA hardware. From these results, a
validity-driven initialization method for AutoTVM is developed, only requiring
41.6% of the necessary hardware measurements to find the best solution, while
improving search robustness. |
3,337 | Few-Shot Unlearning by Model Inversion | We consider the problem of machine unlearning to erase a target dataset,
which causes an unwanted behavior, from the trained model when the training
dataset is not given. Previous works have assumed that the target dataset
indicates all the training data imposing the unwanted behavior. However, it is
often infeasible to obtain such a complete indication. We hence address a
practical scenario of unlearning provided a few samples of target data,
so-called few-shot unlearning. To this end, we devise a straightforward
framework, including a new model inversion technique to retrieve the training
data from the model, followed by filtering out samples similar to the target
samples and then relearning. We demonstrate that our method using only a subset
of target data can outperform the state-of-the-art methods with a full
indication of target data. |
3,338 | Secure Federated Clustering | We consider a foundational unsupervised learning task of $k$-means data
clustering, in a federated learning (FL) setting consisting of a central server
and many distributed clients. We develop SecFC, which is a secure federated
clustering algorithm that simultaneously achieves 1) universal performance: no
performance loss compared with clustering over centralized data, regardless of
data distribution across clients; 2) data privacy: each client's private data
and the cluster centers are not leaked to other clients and the server. In
SecFC, the clients perform Lagrange encoding on their local data and share the
coded data in an information-theoretically private manner; then leveraging the
algebraic structure of the coding, the FL network exactly executes the Lloyd's
$k$-means heuristic over the coded data to obtain the final clustering.
Experiment results on synthetic and real datasets demonstrate the universally
superior performance of SecFC for different data distributions across clients,
and its computational practicality for various combinations of system
parameters. Finally, we propose an extension of SecFC to further provide
membership privacy for all data points. |
3,339 | Graph-level Neural Networks: Current Progress and Future Directions | Graph-structured data consisting of objects (i.e., nodes) and relationships
among objects (i.e., edges) are ubiquitous. Graph-level learning is a matter of
studying a collection of graphs instead of a single graph. Traditional
graph-level learning methods used to be the mainstream. However, with the
increasing scale and complexity of graphs, Graph-level Neural Networks (GLNNs,
deep learning-based graph-level learning methods) have been attractive due to
their superiority in modeling high-dimensional data. Thus, a survey on GLNNs is
necessary. To frame this survey, we propose a systematic taxonomy covering
GLNNs upon deep neural networks, graph neural networks, and graph pooling. The
representative and state-of-the-art models in each category are focused on this
survey. We also investigate the reproducibility, benchmarks, and new graph
datasets of GLNNs. Finally, we conclude future directions to further push
forward GLNNs. The repository of this survey is available at
https://github.com/GeZhangMQ/Awesome-Graph-level-Neural-Networks. |
3,340 | VC Theoretical Explanation of Double Descent | There has been growing interest in generalization performance of large
multilayer neural networks that can be trained to achieve zero training error,
while generalizing well on test data. This regime is known as 'second descent'
and it appears to contradict conventional view that optimal model complexity
should reflect optimal balance between underfitting and overfitting, aka the
bias-variance trade-off. This paper presents VC-theoretical analysis of double
descent and shows that it can be fully explained by classical VC generalization
bounds. We illustrate an application of analytic VC-bounds for modeling double
descent for classification problems, using empirical results for several
learning methods, such as SVM, Least Squares, and Multilayer Perceptron
classifiers. In addition, we discuss several possible reasons for
misinterpretation of VC-theoretical results in the machine learning community. |
3,341 | Robust Projection based Anomaly Extraction (RPE) in Univariate Time-Series | This paper presents a novel, closed-form, and data/computation efficient
online anomaly detection algorithm for time-series data. The proposed method,
dubbed RPE, is a window-based method and in sharp contrast to the existing
window-based methods, it is robust to the presence of anomalies in its window
and it can distinguish the anomalies in time-stamp level. RPE leverages the
linear structure of the trajectory matrix of the time-series and employs a
robust projection step which makes the algorithm able to handle the presence of
multiple arbitrarily large anomalies in its window. A closed-form/non-iterative
algorithm for the robust projection step is provided and it is proved that it
can identify the corrupted time-stamps. RPE is a great candidate for the
applications where a large training data is not available which is the common
scenario in the area of time-series. An extensive set of numerical experiments
show that RPE can outperform the existing approaches with a notable margin. |
3,342 | MACE: An Efficient Model-Agnostic Framework for Counterfactual Explanation | Counterfactual explanation is an important Explainable AI technique to
explain machine learning predictions. Despite being studied actively, existing
optimization-based methods often assume that the underlying machine-learning
model is differentiable and treat categorical attributes as continuous ones,
which restricts their real-world applications when categorical attributes have
many different values or the model is non-differentiable. To make
counterfactual explanation suitable for real-world applications, we propose a
novel framework of Model-Agnostic Counterfactual Explanation (MACE), which
adopts a newly designed pipeline that can efficiently handle non-differentiable
machine-learning models on a large number of feature values. in our MACE
approach, we propose a novel RL-based method for finding good counterfactual
examples and a gradient-less descent method for improving proximity.
Experiments on public datasets validate the effectiveness with better validity,
sparsity and proximity. |
3,343 | DeepDefacer: Automatic Removal of Facial Features via U-Net Image Segmentation | Recent advancements in the field of magnetic resonance imaging (MRI) have
enabled large-scale collaboration among clinicians and researchers for
neuroimaging tasks. However, researchers are often forced to use outdated and
slow software to anonymize MRI images for publication. These programs
specifically perform expensive mathematical operations over 3D images that
rapidly slow down anonymization speed as an image's volume increases in size.
In this paper, we introduce DeepDefacer, an application of deep learning to MRI
anonymization that uses a streamlined 3D U-Net network to mask facial regions
in MRI images with a significant increase in speed over traditional
de-identification software. We train DeepDefacer on MRI images from the Brain
Development Organization (IXI) and International Consortium for Brain Mapping
(ICBM) and quantitatively evaluate our model against a baseline 3D U-Net model
with regards to Dice, recall, and precision scores. We also evaluate
DeepDefacer against Pydeface, a traditional defacing application, with regards
to speed on a range of CPU and GPU devices and qualitatively evaluate our
model's defaced output versus the ground truth images produced by Pydeface. We
provide a link to a PyPi program at the end of this manuscript to encourage
further research into the application of deep learning to MRI anonymization. |
3,344 | Gluing Neural Networks Symbolically Through Hyperdimensional Computing | Hyperdimensional Computing affords simple, yet powerful operations to create
long Hyperdimensional Vectors (hypervectors) that can efficiently encode
information, be used for learning, and are dynamic enough to be modified on the
fly. In this paper, we explore the notion of using binary hypervectors to
directly encode the final, classifying output signals of neural networks in
order to fuse differing networks together at the symbolic level. This allows
multiple neural networks to work together to solve a problem, with little
additional overhead. Output signals just before classification are encoded as
hypervectors and bundled together through consensus summation to train a
classification hypervector. This process can be performed iteratively and even
on single neural networks by instead making a consensus of multiple
classification hypervectors. We find that this outperforms the state of the
art, or is on a par with it, while using very little overhead, as hypervector
operations are extremely fast and efficient in comparison to the neural
networks. This consensus process can learn online and even grow or lose models
in real time. Hypervectors act as memories that can be stored, and even further
bundled together over time, affording life long learning capabilities.
Additionally, this consensus structure inherits the benefits of
Hyperdimensional Computing, without sacrificing the performance of modern
Machine Learning. This technique can be extrapolated to virtually any neural
model, and requires little modification to employ - one simply requires
recording the output signals of networks when presented with a testing example. |
3,345 | itKD: Interchange Transfer-based Knowledge Distillation for 3D Object Detection | Recently, point-cloud based 3D object detectors have achieved remarkable
progress. However, most studies are limited to the development of deep learning
architectures for improving only their accuracy. In this paper, we propose an
autoencoder-style framework comprising channel-wise compression and
decompression via interchange transfer for knowledge distillation. To learn the
map-view feature of a teacher network, the features from a teacher and student
network are independently passed through the shared autoencoder; here, we use a
compressed representation loss that binds the channel-wised compression
knowledge from both the networks as a kind of regularization. The decompressed
features are transferred in opposite directions to reduce the gap in the
interchange reconstructions. Lastly, we present an attentive head loss for
matching the pivotal detection information drawn by the multi-head
self-attention mechanism. Through extensive experiments, we verify that our
method can learn the lightweight model that is well-aligned with the 3D point
cloud detection task and we demonstrate its superiority using the well-known
public datasets Waymo and nuScenes. |
3,346 | Variational Transfer Learning using Cross-Domain Latent Modulation | To successfully apply trained neural network models to new domains, powerful
transfer learning solutions are essential. We propose to introduce a novel
cross-domain latent modulation mechanism to a variational autoencoder framework
so as to achieve effective transfer learning. Our key idea is to procure deep
representations from one data domain and use it to influence the
reparameterization of the latent variable of another domain. Specifically, deep
representations of the source and target domains are first extracted by a
unified inference model and aligned by employing gradient reversal. The learned
deep representations are then cross-modulated to the latent encoding of the
alternative domain, where consistency constraints are also applied. In the
empirical validation that includes a number of transfer learning benchmark
tasks for unsupervised domain adaptation and image-to-image translation, our
model demonstrates competitive performance, which is also supported by evidence
obtained from visualization. |
3,347 | Nearly Minimax Optimal Offline Reinforcement Learning with Linear Function Approximation: Single-Agent MDP and Markov Game | Offline reinforcement learning (RL) aims at learning an optimal strategy
using a pre-collected dataset without further interactions with the
environment. While various algorithms have been proposed for offline RL in the
previous literature, the minimax optimal performance has only been (nearly)
achieved for tabular Markov decision processes (MDPs). In this paper, we focus
on offline RL with linear function approximation and propose two new
algorithms, SPEVI+ and SPMVI+, for single-agent MDPs and two-player zero-sum
Markov games (MGs), respectively. The proposed algorithms feature carefully
crafted data splitting mechanisms and novel variance-reduction pessimistic
estimators. Theoretical analysis demonstrates that they are capable of matching
the performance lower bounds up to logarithmic factors. As a byproduct, a new
performance lower bound is established for MGs, which tightens the existing
results. To the best of our knowledge, these are the first computationally
efficient and nearly minimax optimal algorithms for offline single-agent MDPs
and MGs with linear function approximation. |
3,348 | Molecular Dipole Moment Learning via Rotationally Equivariant Gaussian Process Regression with Derivatives in Molecular-orbital-based Machine Learning | This study extends the accurate and transferable molecular-orbital-based
machine learning (MOB-ML) approach to modeling the contribution of electron
correlation to dipole moments at the cost of Hartree-Fock computations. A
molecular-orbital-based (MOB) pairwise decomposition of the correlation part of
the dipole moment is applied, and these pair dipole moments could be further
regressed as a universal function of molecular orbitals (MOs). The dipole MOB
features consist of the energy MOB features and their responses to electric
fields. An interpretable and rotationally equivariant Gaussian process
regression (GPR) with derivatives algorithm is introduced to learn the dipole
moment more efficiently. The proposed problem setup, feature design, and ML
algorithm are shown to provide highly-accurate models for both dipole moment
and energies on water and fourteen small molecules. To demonstrate the ability
of MOB-ML to function as generalized density-matrix functionals for molecular
dipole moments and energies of organic molecules, we further apply the proposed
MOB-ML approach to train and test the molecules from the QM9 dataset. The
application of local scalable GPR with Gaussian mixture model unsupervised
clustering (GMM/GPR) scales up MOB-ML to a large-data regime while retaining
the prediction accuracy. In addition, compared with literature results, MOB-ML
provides the best test MAEs of 4.21 mDebye and 0.045 kcal/mol for dipole moment
and energy models, respectively, when training on 110000 QM9 molecules. The
excellent transferability of the resulting QM9 models is also illustrated by
the accurate predictions for four different series of peptides. |
3,349 | Rethinking Graph Neural Networks for Anomaly Detection | Graph Neural Networks (GNNs) are widely applied for graph anomaly detection.
As one of the key components for GNN design is to select a tailored spectral
filter, we take the first step towards analyzing anomalies via the lens of the
graph spectrum. Our crucial observation is the existence of anomalies will lead
to the `right-shift' phenomenon, that is, the spectral energy distribution
concentrates less on low frequencies and more on high frequencies. This fact
motivates us to propose the Beta Wavelet Graph Neural Network (BWGNN). Indeed,
BWGNN has spectral and spatial localized band-pass filters to better handle the
`right-shift' phenomenon in anomalies. We demonstrate the effectiveness of
BWGNN on four large-scale anomaly detection datasets. Our code and data are
released at https://github.com/squareRoot3/Rethinking-Anomaly-Detection |
3,350 | Certifying Some Distributional Fairness with Subpopulation Decomposition | Extensive efforts have been made to understand and improve the fairness of
machine learning models based on observational metrics, especially in
high-stakes domains such as medical insurance, education, and hiring decisions.
However, there is a lack of certified fairness considering the end-to-end
performance of an ML model. In this paper, we first formulate the certified
fairness of an ML model trained on a given data distribution as an optimization
problem based on the model performance loss bound on a fairness constrained
distribution, which is within bounded distributional distance with the training
distribution. We then propose a general fairness certification framework and
instantiate it for both sensitive shifting and general shifting scenarios. In
particular, we propose to solve the optimization problem by decomposing the
original data distribution into analytical subpopulations and proving the
convexity of the subproblems to solve them. We evaluate our certified fairness
on six real-world datasets and show that our certification is tight in the
sensitive shifting scenario and provides non-trivial certification under
general shifting. Our framework is flexible to integrate additional
non-skewness constraints and we show that it provides even tighter
certification under different real-world scenarios. We also compare our
certified fairness bound with adapted existing distributional robustness bounds
on Gaussian data and demonstrate that our method is significantly tighter. |
3,351 | Sepsis Prediction with Temporal Convolutional Networks | We design and implement a temporal convolutional network model to predict
sepsis onset. Our model is trained on data extracted from MIMIC III database,
based on a retrospective analysis of patients admitted to intensive care unit
who did not fall under the definition of sepsis at the time of admission.
Benchmarked with several machine learning models, our model is superior on this
binary classification task, demonstrates the prediction power of convolutional
networks for temporal patterns, also shows the significant impact of having
longer look back time on sepsis prediction. |
3,352 | Post-hoc Concept Bottleneck Models | Concept Bottleneck Models (CBMs) map the inputs onto a set of interpretable
concepts (``the bottleneck'') and use the concepts to make predictions. A
concept bottleneck enhances interpretability since it can be investigated to
understand what concepts the model "sees" in an input and which of these
concepts are deemed important. However, CBMs are restrictive in practice as
they require concept labels in the training data to learn the bottleneck and do
not leverage strong pretrained models. Moreover, CBMs often do not match the
accuracy of an unrestricted neural network, reducing the incentive to deploy
them in practice. In this work, we address the limitations of CBMs by
introducing Post-hoc Concept Bottleneck models (PCBMs). We show that we can
turn any neural network into a PCBM without sacrificing model performance while
still retaining interpretability benefits. When concept annotation is not
available on the training data, we show that PCBM can transfer concepts from
other datasets or from natural language descriptions of concepts. PCBM also
enables users to quickly debug and update the model to reduce spurious
correlations and improve generalization to new (potentially different) data.
Through a model-editing user study, we show that editing PCBMs via
concept-level feedback can provide significant performance gains without using
any data from the target domain or model retraining. |
3,353 | Data Banzhaf: A Data Valuation Framework with Maximal Robustness to Learning Stochasticity | This paper studies the robustness of data valuation to noisy model
performance scores. Particularly, we find that the inherent randomness of the
widely used stochastic gradient descent can cause existing data value notions
(e.g., the Shapley value and the Leave-one-out error) to produce inconsistent
data value rankings across different runs. To address this challenge, we first
pose a formal framework within which one can measure the robustness of a data
value notion. We show that the Banzhaf value, a value notion originated from
cooperative game theory literature, achieves the maximal robustness among all
semivalues -- a class of value notions that satisfy crucial properties entailed
by ML applications. We propose an algorithm to efficiently estimate the Banzhaf
value based on the Maximum Sample Reuse (MSR) principle. We derive the lower
bound sample complexity for Banzhaf value approximation, and we show that our
MSR algorithm's sample complexity nearly matches the lower bound. Our
evaluation demonstrates that the Banzhaf value outperforms the existing
semivalue-based data value notions on several downstream ML tasks such as
learning with weighted samples and noisy label detection. Overall, our study
suggests that when the underlying ML algorithm is stochastic, the Banzhaf value
is a promising alternative to the semivalue-based data value schemes given its
computational advantage and ability to robustly differentiate data quality. |
3,354 | Few-Shot Diffusion Models | Denoising diffusion probabilistic models (DDPM) are powerful hierarchical
latent variable models with remarkable sample generation quality and training
stability. These properties can be attributed to parameter sharing in the
generative hierarchy, as well as a parameter-free diffusion-based inference
procedure. In this paper, we present Few-Shot Diffusion Models (FSDM), a
framework for few-shot generation leveraging conditional DDPMs. FSDMs are
trained to adapt the generative process conditioned on a small set of images
from a given class by aggregating image patch information using a set-based
Vision Transformer (ViT). At test time, the model is able to generate samples
from previously unseen classes conditioned on as few as 5 samples from that
class. We empirically show that FSDM can perform few-shot generation and
transfer to new datasets. We benchmark variants of our method on complex vision
datasets for few-shot learning and compare to unconditional and conditional
DDPM baselines. Additionally, we show how conditioning the model on patch-based
input set information improves training convergence. |
3,355 | Critic Sequential Monte Carlo | We introduce CriticSMC, a new algorithm for planning as inference built from
a novel composition of sequential Monte Carlo with learned soft-Q function
heuristic factors. This algorithm is structured so as to allow using large
numbers of putative particles leading to efficient utilization of computational
resource and effective discovery of high reward trajectories even in
environments with difficult reward surfaces such as those arising from hard
constraints. Relative to prior art our approach is notably still compatible
with model-free reinforcement learning in the sense that the implicit policy we
produce can be used at test time in the absence of a world model. Our
experiments on self-driving car collision avoidance in simulation demonstrate
improvements against baselines in terms of infraction minimization relative to
computational effort while maintaining diversity and realism of found
trajectories. |
3,356 | A Comparative Study on Energy Consumption Models for Drones | Creating an appropriate energy consumption prediction model is becoming an
important topic for drone-related research in the literature. However, a
general consensus on the energy consumption model is yet to be reached at
present. As a result, there are many variations that attempt to create models
that range in complexity with a focus on different aspects. In this paper, we
benchmark the five most popular energy consumption models for drones derived
from their physical behaviours and point to the difficulties in matching with a
realistic energy dataset collected from a delivery drone in flight under
different testing conditions. Moreover, we propose a novel data-driven energy
model using the Long Short-Term Memory (LSTM) based deep learning architecture
and the accuracy is compared based on the dataset. Our experimental results
have shown that the LSTM based approach can easily outperform other
mathematical models for the dataset under study. Finally, sensitivity analysis
has been carried out in order to interpret the model. |
3,357 | Bayesian Active Learning for Scanning Probe Microscopy: from Gaussian Processes to Hypothesis Learning | Recent progress in machine learning methods, and the emerging availability of
programmable interfaces for scanning probe microscopes (SPMs), have propelled
automated and autonomous microscopies to the forefront of attention of the
scientific community. However, enabling automated microscopy requires the
development of task-specific machine learning methods, understanding the
interplay between physics discovery and machine learning, and fully defined
discovery workflows. This, in turn, requires balancing the physical intuition
and prior knowledge of the domain scientist with rewards that define
experimental goals and machine learning algorithms that can translate these to
specific experimental protocols. Here, we discuss the basic principles of
Bayesian active learning and illustrate its applications for SPM. We progress
from the Gaussian Process as a simple data-driven method and Bayesian inference
for physical models as an extension of physics-based functional fits to more
complex deep kernel learning methods, structured Gaussian Processes, and
hypothesis learning. These frameworks allow for the use of prior data, the
discovery of specific functionalities as encoded in spectral data, and
exploration of physical laws manifesting during the experiment. The discussed
framework can be universally applied to all techniques combining imaging and
spectroscopy, SPM methods, nanoindentation, electron microscopy and
spectroscopy, and chemical imaging methods, and can be particularly impactful
for destructive or irreversible measurements. |
3,358 | GLDQN: Explicitly Parameterized Quantile Reinforcement Learning for Waste Reduction | We study the problem of restocking a grocery store's inventory with
perishable items over time, from a distributional point of view. The objective
is to maximize sales while minimizing waste, with uncertainty about the actual
consumption by costumers. This problem is of a high relevance today, given the
growing demand for food and the impact of food waste on the environment, the
economy, and purchasing power. We frame inventory restocking as a new
reinforcement learning task that exhibits stochastic behavior conditioned on
the agent's actions, making the environment partially observable. We introduce
a new reinforcement learning environment based on real grocery store data and
expert knowledge. This environment is highly stochastic, and presents a unique
challenge for reinforcement learning practitioners. We show that uncertainty
about the future behavior of the environment is not handled well by classical
supply chain algorithms, and that distributional approaches are a good way to
account for the uncertainty. We also present GLDQN, a new distributional
reinforcement learning algorithm that learns a generalized lambda distribution
over the reward space. We show that GLDQN outperforms other distributional
reinforcement learning approaches in our partially observable environments, in
both overall reward and generated waste. |
3,359 | Posterior and Computational Uncertainty in Gaussian Processes | Gaussian processes scale prohibitively with the size of the dataset. In
response, many approximation methods have been developed, which inevitably
introduce approximation error. This additional source of uncertainty, due to
limited computation, is entirely ignored when using the approximate posterior.
Therefore in practice, GP models are often as much about the approximation
method as they are about the data. Here, we develop a new class of methods that
provides consistent estimation of the combined uncertainty arising from both
the finite number of data observed and the finite amount of computation
expended. The most common GP approximations map to an instance in this class,
such as methods based on the Cholesky factorization, conjugate gradients, and
inducing points. For any method in this class, we prove (i) convergence of its
posterior mean in the associated RKHS, (ii) decomposability of its combined
posterior covariance into mathematical and computational covariances, and (iii)
that the combined variance is a tight worst-case bound for the squared error
between the method's posterior mean and the latent function. Finally, we
empirically demonstrate the consequences of ignoring computational uncertainty
and show how implicitly modeling it improves generalization performance on
benchmark datasets. |
3,360 | Holistic Generalized Linear Models | Holistic linear regression extends the classical best subset selection
problem by adding additional constraints designed to improve the model quality.
These constraints include sparsity-inducing constraints, sign-coherence
constraints and linear constraints. The $\textsf{R}$ package $\texttt{holiglm}$
provides functionality to model and fit holistic generalized linear models. By
making use of state-of-the-art conic mixed-integer solvers, the package can
reliably solve GLMs for Gaussian, binomial and Poisson responses with a
multitude of holistic constraints. The high-level interface simplifies the
constraint specification and can be used as a drop-in replacement for the
$\texttt{stats::glm()}$ function. |
3,361 | Continual Object Detection: A review of definitions, strategies, and challenges | The field of Continual Learning investigates the ability to learn consecutive
tasks without losing performance on those previously learned. Its focus has
been mainly on incremental classification tasks. We believe that research in
continual object detection deserves even more attention due to its vast range
of applications in robotics and autonomous vehicles. This scenario is more
complex than conventional classification given the occurrence of instances of
classes that are unknown at the time, but can appear in subsequent tasks as a
new class to be learned, resulting in missing annotations and conflicts with
the background label. In this review, we analyze the current strategies
proposed to tackle the problem of class-incremental object detection. Our main
contributions are: (1) a short and systematic review of the methods that
propose solutions to traditional incremental object detection scenarios; (2) A
comprehensive evaluation of the existing approaches using a new metric to
quantify the stability and plasticity of each technique in a standard way; (3)
an overview of the current trends within continual object detection and a
discussion of possible future research directions. |
3,362 | StyleTTS: A Style-Based Generative Model for Natural and Diverse Text-to-Speech Synthesis | Text-to-Speech (TTS) has recently seen great progress in synthesizing
high-quality speech owing to the rapid development of parallel TTS systems, but
producing speech with naturalistic prosodic variations, speaking styles and
emotional tones remains challenging. Moreover, since duration and speech are
generated separately, parallel TTS models still have problems finding the best
monotonic alignments that are crucial for naturalistic speech synthesis. Here,
we propose StyleTTS, a style-based generative model for parallel TTS that can
synthesize diverse speech with natural prosody from a reference speech
utterance. With novel Transferable Monotonic Aligner (TMA) and
duration-invariant data augmentation schemes, our method significantly
outperforms state-of-the-art models on both single and multi-speaker datasets
in subjective tests of speech naturalness and speaker similarity. Through
self-supervised learning of the speaking styles, our model can synthesize
speech with the same prosodic and emotional tone as any given reference speech
without the need for explicitly labeling these categories. |
3,363 | FBM: Fast-Bit Allocation for Mixed-Precision Quantization | Quantized neural networks are well known for reducing latency, power
consumption, and model size without significant degradation in accuracy, making
them highly applicable for systems with limited resources and low power
requirements.
Mixed precision quantization offers better utilization of customized hardware
that supports arithmetic operations at different bitwidths. Existing
mixed-precision schemes rely on having a high exploration space, resulting in a
large carbon footprint. In addition, these bit allocation strategies mostly
induce constraints on the model size rather than utilizing the performance of
neural network deployment on specific hardware. Our work proposes Fast-Bit
Allocation for Mixed-Precision Quantization (FBM), which finds an optimal
bitwidth allocation by measuring desired behaviors through a simulation of a
specific device, or even on a physical one.
While dynamic transitions of bit allocation in mixed precision quantization
with ultra-low bitwidth are known to suffer from performance degradation, we
present a fast recovery solution from such transitions.
A comprehensive evaluation of the proposed method on CIFAR-10 and ImageNet
demonstrates our method's superiority over current state-of-the-art schemes in
terms of the trade-off between neural network accuracy and hardware efficiency.
Our source code, experimental settings and quantized models are available at
https://github.com/RamorayDrake/FBM/ |
3,364 | Fairness in the First Stage of Two-Stage Recommender Systems | Many large-scale recommender systems consist of two stages, where the first
stage focuses on efficiently generating a small subset of promising candidates
from a huge pool of items for the second-stage model to curate final
recommendations from. In this paper, we investigate how to ensure group
fairness to the items in this two-stage paradigm. In particular, we find that
existing first-stage recommenders might select an irrecoverably unfair set of
candidates such that there is no hope for the second-stage recommender to
deliver fair recommendations. To this end, we propose two threshold-policy
selection rules that, given any relevance model of queries and items and a
point-wise lower confidence bound on the expected number of relevant items for
each policy, find near-optimal sets of candidates that contain enough relevant
items in expectation from each group of items. To instantiate the rules, we
demonstrate how to derive such confidence bounds from potentially partial and
biased user feedback data, which are abundant in many large-scale recommender
systems. In addition, we provide both finite-sample and asymptotic analysis of
how close the two threshold selection rules are to the optimal thresholds.
Beyond this theoretical analysis, we show empirically that these two rules can
consistently select enough relevant items from each group while minimizing the
size of the candidate sets for a wide range of settings. |
3,365 | Learning Risk-Averse Equilibria in Multi-Agent Systems | In multi-agent systems, intelligent agents are tasked with making decisions
that have optimal outcomes when the actions of the other agents are as
expected, whilst also being prepared for unexpected behaviour. In this work, we
introduce a new risk-averse solution concept that allows the learner to
accommodate unexpected actions by finding the minimum variance strategy given
any level of expected return. We prove the existence of such a risk-averse
equilibrium, and propose one fictitious-play type learning algorithm for
smaller games that enjoys provable convergence guarantees in certain games
classes (e.g., zero-sum or potential). Furthermore, we propose an approximation
method for larger games based on iterative population-based training that
generates a population of risk-averse agents. Empirically, our equilibrium is
shown to be able to reduce the reward variance, specifically in the sense that
off-equilibrium behaviour has a far smaller impact on our risk-averse agents in
comparison to playing other equilibrium solutions. Importantly, we show that
our population of agents that approximate a risk-averse equilibrium is
particularly effective in the presence of unseen opposing populations,
especially in the case of guaranteeing a minimal level of performance which is
critical to safety-aware multi-agent systems. |
3,366 | Towards Fair Federated Recommendation Learning: Characterizing the Inter-Dependence of System and Data Heterogeneity | Federated learning (FL) is an effective mechanism for data privacy in
recommender systems by running machine learning model training on-device. While
prior FL optimizations tackled the data and system heterogeneity challenges
faced by FL, they assume the two are independent of each other. This
fundamental assumption is not reflective of real-world, large-scale recommender
systems -- data and system heterogeneity are tightly intertwined. This paper
takes a data-driven approach to show the inter-dependence of data and system
heterogeneity in real-world data and quantifies its impact on the overall model
quality and fairness. We design a framework, RF^2, to model the
inter-dependence and evaluate its impact on state-of-the-art model optimization
techniques for federated recommendation tasks. We demonstrate that the impact
on fairness can be severe under realistic heterogeneity scenarios, by up to
15.8--41x compared to a simple setup assumed in most (if not all) prior work.
It means when realistic system-induced data heterogeneity is not properly
modeled, the fairness impact of an optimization can be downplayed by up to 41x.
The result shows that modeling realistic system-induced data heterogeneity is
essential to achieving fair federated recommendation learning. We plan to
open-source RF^2 to enable future design and evaluation of FL innovations. |
3,367 | Segmentation Consistency Training: Out-of-Distribution Generalization for Medical Image Segmentation | Generalizability is seen as one of the major challenges in deep learning, in
particular in the domain of medical imaging, where a change of hospital or in
imaging routines can lead to a complete failure of a model. To tackle this, we
introduce Consistency Training, a training procedure and alternative to data
augmentation based on maximizing models' prediction consistency across
augmented and unaugmented data in order to facilitate better
out-of-distribution generalization. To this end, we develop a novel
region-based segmentation loss function called Segmentation Inconsistency Loss
(SIL), which considers the differences between pairs of augmented and
unaugmented predictions and labels. We demonstrate that Consistency Training
outperforms conventional data augmentation on several out-of-distribution
datasets on polyp segmentation, a popular medical task. |
3,368 | Connecting adversarial attacks and optimal transport for domain adaptation | We present a novel algorithm for domain adaptation using optimal transport.
In domain adaptation, the goal is to adapt a classifier trained on the source
domain samples to the target domain. In our method, we use optimal transport to
map target samples to the domain named source fiction. This domain differs from
the source but is accurately classified by the source domain classifier. Our
main idea is to generate a source fiction by c-cyclically monotone
transformation over the target domain. If samples with the same labels in two
domains are c-cyclically monotone, the optimal transport map between these
domains preserves the class-wise structure, which is the main goal of domain
adaptation. To generate a source fiction domain, we propose an algorithm that
is based on our finding that adversarial attacks are a c-cyclically monotone
transformation of the dataset. We conduct experiments on Digits and Modern
Office-31 datasets and achieve improvement in performance for simple discrete
optimal transport solvers for all adaptation tasks. |
3,369 | Analysis of Augmentations for Contrastive ECG Representation Learning | This paper systematically investigates the effectiveness of various
augmentations for contrastive self-supervised learning of electrocardiogram
(ECG) signals and identifies the best parameters. The baseline of our proposed
self-supervised framework consists of two main parts: the contrastive learning
and the downstream task. In the first stage, we train an encoder using a number
of augmentations to extract generalizable ECG signal representations. We then
freeze the encoder and finetune a few linear layers with different amounts of
labelled data for downstream arrhythmia detection. We then experiment with
various augmentations techniques and explore a range of parameters. Our
experiments are done on PTB-XL, a large and publicly available 12-lead ECG
dataset. The results show that applying augmentations in a specific range of
complexities works better for self-supervised contrastive learning. For
instance, when adding Gaussian noise, a sigma in the range of 0.1 to 0.2
achieves better results, while poor training occurs when the added noise is too
small or too large (outside of the specified range). A similar trend is
observed with other augmentations, demonstrating the importance of selecting
the optimum level of difficulty for the added augmentations, as augmentations
that are too simple will not result in effective training, while augmentations
that are too difficult will also prevent the model from effective learning of
generalized representations. Our work can influence future research on
self-supervised contrastive learning on bio-signals and aid in selecting
optimum parameters for different augmentations. |
3,370 | Fooling SHAP with Stealthily Biased Sampling | SHAP explanations aim at identifying which features contribute the most to
the difference in model prediction at a specific input versus a background
distribution. Recent studies have shown that they can be manipulated by
malicious adversaries to produce arbitrary desired explanations. However,
existing attacks focus solely on altering the black-box model itself. In this
paper, we propose a complementary family of attacks that leave the model intact
and manipulate SHAP explanations using stealthily biased sampling of the data
points used to approximate expectations w.r.t the background distribution. In
the context of fairness audit, we show that our attack can reduce the
importance of a sensitive feature when explaining the difference in outcomes
between groups, while remaining undetected. These results highlight the
manipulability of SHAP explanations and encourage auditors to treat post-hoc
explanations with skepticism. |
3,371 | PolypConnect: Image inpainting for generating realistic gastrointestinal tract images with polyps | Early identification of a polyp in the lower gastrointestinal (GI) tract can
lead to prevention of life-threatening colorectal cancer. Developing
computer-aided diagnosis (CAD) systems to detect polyps can improve detection
accuracy and efficiency and save the time of the domain experts called
endoscopists. Lack of annotated data is a common challenge when building CAD
systems. Generating synthetic medical data is an active research area to
overcome the problem of having relatively few true positive cases in the
medical domain. To be able to efficiently train machine learning (ML) models,
which are the core of CAD systems, a considerable amount of data should be
used. In this respect, we propose the PolypConnect pipeline, which can convert
non-polyp images into polyp images to increase the size of training datasets
for training. We present the whole pipeline with quantitative and qualitative
evaluations involving endoscopists. The polyp segmentation model trained using
synthetic data, and real data shows a 5.1% improvement of mean intersection
over union (mIOU), compared to the model trained only using real data. The
codes of all the experiments are available on GitHub to reproduce the results. |
3,372 | Grid HTM: Hierarchical Temporal Memory for Anomaly Detection in Videos | The interest for video anomaly detection systems has gained traction for the
past few years. The current approaches use deep learning to perform anomaly
detection in videos, but this approach has multiple problems. For starters,
deep learning in general has issues with noise, concept drift, explainability,
and training data volumes. Additionally, anomaly detection in itself is a
complex task and faces challenges such as unknowness, heterogeneity, and class
imbalance. Anomaly detection using deep learning is therefore mainly
constrained to generative models such as generative adversarial networks and
autoencoders due to their unsupervised nature, but even they suffer from
general deep learning issues and are hard to train properly. In this paper, we
explore the capabilities of the Hierarchical Temporal Memory (HTM) algorithm to
perform anomaly detection in videos, as it has favorable properties such as
noise tolerance and online learning which combats concept drift. We introduce a
novel version of HTM, namely, Grid HTM, which is an HTM-based architecture
specifically for anomaly detection in complex videos such as surveillance
footage. |
3,373 | Neural Optimal Transport with General Cost Functionals | We present a novel neural-networks-based algorithm to compute optimal
transport (OT) plans and maps for general cost functionals. The algorithm is
based on a saddle point reformulation of the OT problem and generalizes prior
OT methods for weak and strong cost functionals. As an application, we
construct a functional to map data distributions with preserving the class-wise
structure of data. |
3,374 | Designing Rewards for Fast Learning | To convey desired behavior to a Reinforcement Learning (RL) agent, a designer
must choose a reward function for the environment, arguably the most important
knob designers have in interacting with RL agents. Although many reward
functions induce the same optimal behavior (Ng et al., 1999), in practice, some
of them result in faster learning than others. In this paper, we look at how
reward-design choices impact learning speed and seek to identify principles of
good reward design that quickly induce target behavior. This
reward-identification problem is framed as an optimization problem: Firstly, we
advocate choosing state-based rewards that maximize the action gap, making
optimal actions easy to distinguish from suboptimal ones. Secondly, we propose
minimizing a measure of the horizon, something we call the "subjective
discount", over which rewards need to be optimized to encourage agents to make
optimal decisions with less lookahead. To solve this optimization problem, we
propose a linear-programming based algorithm that efficiently finds a reward
function that maximizes action gap and minimizes subjective discount. We test
the rewards generated with the algorithm in tabular environments with
Q-Learning, and empirically show they lead to faster learning. Although we only
focus on Q-Learning because it is perhaps the simplest and most well understood
RL algorithm, preliminary results with R-max (Brafman and Tennenholtz, 2000)
suggest our results are much more general. Our experiments support three
principles of reward design: 1) consistent with existing results, penalizing
each step taken induces faster learning than rewarding the goal. 2) When
rewarding subgoals along the target trajectory, rewards should gradually
increase as the goal gets closer. 3) Dense reward that's nonzero on every state
is only good if designed carefully. |
3,375 | Minimax Optimal Online Imitation Learning via Replay Estimation | Online imitation learning is the problem of how best to mimic expert
demonstrations, given access to the environment or an accurate simulator. Prior
work has shown that in the infinite sample regime, exact moment matching
achieves value equivalence to the expert policy. However, in the finite sample
regime, even if one has no optimization error, empirical variance can lead to a
performance gap that scales with $H^2 / N$ for behavioral cloning and $H /
\sqrt{N}$ for online moment matching, where $H$ is the horizon and $N$ is the
size of the expert dataset. We introduce the technique of replay estimation to
reduce this empirical variance: by repeatedly executing cached expert actions
in a stochastic simulator, we compute a smoother expert visitation distribution
estimate to match. In the presence of general function approximation, we prove
a meta theorem reducing the performance gap of our approach to the parameter
estimation error for offline classification (i.e. learning the expert policy).
In the tabular setting or with linear function approximation, our meta theorem
shows that the performance gap incurred by our approach achieves the optimal
$\widetilde{O} \left( \min({H^{3/2}} / {N}, {H} / {\sqrt{N}} \right)$
dependency, under significantly weaker assumptions compared to prior work. We
implement multiple instantiations of our approach on several continuous control
tasks and find that we are able to significantly improve policy performance
across a variety of dataset sizes. |
3,376 | A multimodal model with Twitter FinBERT embeddings for extreme price movement prediction of Bitcoin | Bitcoin, with its ever-growing popularity, has demonstrated extreme price
volatility since its origin. This volatility, together with its decentralised
nature, make Bitcoin highly subjective to speculative trading as compared to
more traditional assets. In this paper, we propose a multimodal model for
predicting extreme price fluctuations. This model takes as input a variety of
correlated assets, technical indicators, as well as Twitter content. In an
in-depth study, we explore whether social media discussions from the general
public on Bitcoin have predictive power for extreme price movements. A dataset
of 5,000 tweets per day containing the keyword `Bitcoin' was collected from
2015 to 2021. This dataset, called PreBit, is made available online. In our
hybrid model, we use sentence-level FinBERT embeddings, pretrained on financial
lexicons, so as to capture the full contents of the tweets and feed it to the
model in an understandable way. By combining these embeddings with a
Convolutional Neural Network, we built a predictive model for significant
market movements. The final multimodal ensemble model includes this NLP model
together with a model based on candlestick data, technical indicators and
correlated asset prices. In an ablation study, we explore the contribution of
the individual modalities. Finally, we propose and backtest a trading strategy
based on the predictions of our models with varying prediction threshold and
show that it can used to build a profitable trading strategy with a reduced
risk over a `hold' or moving average strategy. |
3,377 | A hybrid approach to seismic deblending: when physics meets self-supervision | To limit the time, cost, and environmental impact associated with the
acquisition of seismic data, in recent decades considerable effort has been put
into so-called simultaneous shooting acquisitions, where seismic sources are
fired at short time intervals between each other. As a consequence, waves
originating from consecutive shots are entangled within the seismic recordings,
yielding so-called blended data. For processing and imaging purposes, the data
generated by each individual shot must be retrieved. This process, called
deblending, is achieved by solving an inverse problem which is heavily
underdetermined. Conventional approaches rely on transformations that render
the blending noise into burst-like noise, whilst preserving the signal of
interest. Compressed sensing type regularization is then applied, where
sparsity in some domain is assumed for the signal of interest. The domain of
choice depends on the geometry of the acquisition and the properties of seismic
data within the chosen domain. In this work, we introduce a new concept that
consists of embedding a self-supervised denoising network into the
Plug-and-Play (PnP) framework. A novel network is introduced whose design
extends the blind-spot network architecture of [28 ] for partially coherent
noise (i.e., correlated in time). The network is then trained directly on the
noisy input data at each step of the PnP algorithm. By leveraging both the
underlying physics of the problem and the great denoising capabilities of our
blind-spot network, the proposed algorithm is shown to outperform an
industry-standard method whilst being comparable in terms of computational
cost. Moreover, being independent on the acquisition geometry, our method can
be easily applied to both marine and land data without any significant
modification. |
3,378 | Attention Flows for General Transformers | In this paper, we study the computation of how much an input token in a
Transformer model influences its prediction. We formalize a method to construct
a flow network out of the attention values of encoder-only Transformer models
and extend it to general Transformer architectures including an auto-regressive
decoder. We show that running a maxflow algorithm on the flow network
construction yields Shapley values, which determine the impact of a player in
cooperative game theory. By interpreting the input tokens in the flow network
as players, we can compute their influence on the total attention flow leading
to the decoder's decision. Additionally, we provide a library that computes and
visualizes the attention flow of arbitrary Transformer models. We show the
usefulness of our implementation on various models trained on natural language
processing and reasoning tasks. |
3,379 | Truly Deterministic Policy Optimization | In this paper, we present a policy gradient method that avoids exploratory
noise injection and performs policy search over the deterministic landscape. By
avoiding noise injection all sources of estimation variance can be eliminated
in systems with deterministic dynamics (up to the initial state distribution).
Since deterministic policy regularization is impossible using traditional
non-metric measures such as the KL divergence, we derive a Wasserstein-based
quadratic model for our purposes. We state conditions on the system model under
which it is possible to establish a monotonic policy improvement guarantee,
propose a surrogate function for policy gradient estimation, and show that it
is possible to compute exact advantage estimates if both the state transition
model and the policy are deterministic. Finally, we describe two novel robotic
control environments -- one with non-local rewards in the frequency domain and
the other with a long horizon (8000 time-steps) -- for which our policy
gradient method (TDPO) significantly outperforms existing methods (PPO, TRPO,
DDPG, and TD3). Our implementation with all the experimental settings is
available at https://github.com/ehsansaleh/code_tdpo |
3,380 | Reinforcement Learning with a Terminator | We present the problem of reinforcement learning with exogenous termination.
We define the Termination Markov Decision Process (TerMDP), an extension of the
MDP framework, in which episodes may be interrupted by an external
non-Markovian observer. This formulation accounts for numerous real-world
situations, such as a human interrupting an autonomous driving agent for
reasons of discomfort. We learn the parameters of the TerMDP and leverage the
structure of the estimation problem to provide state-wise confidence bounds. We
use these to construct a provably-efficient algorithm, which accounts for
termination, and bound its regret. Motivated by our theoretical analysis, we
design and implement a scalable approach, which combines optimism (w.r.t.
termination) and a dynamic discount factor, incorporating the termination
probability. We deploy our method on high-dimensional driving and MinAtar
benchmarks. Additionally, we test our approach on human data in a driving
setting. Our results demonstrate fast convergence and significant improvement
over various baseline approaches. |
3,381 | Optimistic Whittle Index Policy: Online Learning for Restless Bandits | Restless multi-armed bandits (RMABs) extend multi-armed bandits to allow for
stateful arms, where the state of each arm evolves restlessly with different
transitions depending on whether that arm is pulled. However, solving RMABs
requires information on transition dynamics, which is often not available
upfront. To plan in RMAB settings with unknown transitions, we propose the
first online learning algorithm based on the Whittle index policy, using an
upper confidence bound (UCB) approach to learn transition dynamics.
Specifically, we formulate a bilinear program to compute the optimistic Whittle
index from the confidence bounds in transition dynamics. Our algorithm,
UCWhittle, achieves sublinear $O(\sqrt{T \log T})$ frequentist regret to solve
RMABs with unknown transitions. Empirically, we demonstrate that UCWhittle
leverages the structure of RMABs and the Whittle index policy solution to
achieve better performance than existing online learning baselines across three
domains, including on real-world maternal and childcare data aimed at reducing
maternal mortality. |
3,382 | Infinite-dimensional optimization and Bayesian nonparametric learning of stochastic differential equations | The paper has two major themes. The first part of the paper establishes
certain general results for infinite-dimensional optimization problems on
Hilbert spaces. These results cover the classical representer theorem and many
of its variants as special cases and offer a wider scope of applications. The
second part of the paper then develops a systematic approach for learning the
drift function of a stochastic differential equation by integrating the results
of the first part with Bayesian hierarchical framework. Importantly, our
Baysian approach incorporates low-cost sparse learning through proper use of
shrinkage priors while allowing proper quantification of uncertainty through
posterior distributions. Several examples at the end illustrate the accuracy of
our learning scheme. |
3,383 | Non-Markovian Reward Modelling from Trajectory Labels via Interpretable Multiple Instance Learning | We generalise the problem of reward modelling (RM) for reinforcement learning
(RL) to handle non-Markovian rewards. Existing work assumes that human
evaluators observe each step in a trajectory independently when providing
feedback on agent behaviour. In this work, we remove this assumption, extending
RM to include hidden state information that captures temporal dependencies in
human assessment of trajectories. We then show how RM can be approached as a
multiple instance learning (MIL) problem, and develop new MIL models that are
able to capture the time dependencies in labelled trajectories. We demonstrate
on a range of RL tasks that our novel MIL models can reconstruct reward
functions to a high level of accuracy, and that they provide interpretable
learnt hidden information that can be used to train high-performing agent
policies. |
3,384 | Searching for the Essence of Adversarial Perturbations | Neural networks have achieved the state-of-the-art performance on various
machine learning fields, yet the incorporation of malicious perturbations with
input data (adversarial example) is able to fool neural networks' predictions.
This would lead to potential risks in real-world applications, for example,
auto piloting and facial recognition. However, the reason for the existence of
adversarial examples remains controversial. Here we demonstrate that
adversarial perturbations contain human-recognizable information, which is the
key conspirator responsible for a neural network's erroneous prediction. This
concept of human-recognizable information allows us to explain key features
related to adversarial perturbations, which include the existence of
adversarial examples, the transferability among different neural networks, and
the increased neural network interpretability for adversarial training. Two
unique properties in adversarial perturbations that fool neural networks are
uncovered: masking and generation. A special class, the complementary class, is
identified when neural networks classify input images. The human-recognizable
information contained in adversarial perturbations allows researchers to gain
insight on the working principles of neural networks and may lead to develop
techniques that detect/defense adversarial attacks. |
3,385 | Adapting Rapid Motor Adaptation for Bipedal Robots | Recent advances in legged locomotion have enabled quadrupeds to walk on
challenging terrains. However, bipedal robots are inherently more unstable and
hence it's harder to design walking controllers for them. In this work, we
leverage recent advances in rapid adaptation for locomotion control, and extend
them to work on bipedal robots. Similar to existing works, we start with a base
policy which produces actions while taking as input an estimated extrinsics
vector from an adaptation module. This extrinsics vector contains information
about the environment and enables the walking controller to rapidly adapt
online. However, the extrinsics estimator could be imperfect, which might lead
to poor performance of the base policy which expects a perfect estimator. In
this paper, we propose A-RMA (Adapting RMA), which additionally adapts the base
policy for the imperfect extrinsics estimator by finetuning it using model-free
RL. We demonstrate that A-RMA outperforms a number of RL-based baseline
controllers and model-based controllers in simulation, and show zero-shot
deployment of a single A-RMA policy to enable a bipedal robot, Cassie, to walk
in a variety of different scenarios in the real world beyond what it has seen
during training. Videos and results at https://ashish-kmr.github.io/a-rma/ |
3,386 | Efficient $Φ$-Regret Minimization in Extensive-Form Games via Online Mirror Descent | A conceptually appealing approach for learning Extensive-Form Games (EFGs) is
to convert them to Normal-Form Games (NFGs). This approach enables us to
directly translate state-of-the-art techniques and analyses in NFGs to learning
EFGs, but typically suffers from computational intractability due to the
exponential blow-up of the game size introduced by the conversion. In this
paper, we address this problem in natural and important setups for the
\emph{$\Phi$-Hedge} algorithm -- A generic algorithm capable of learning a
large class of equilibria for NFGs. We show that $\Phi$-Hedge can be directly
used to learn Nash Equilibria (zero-sum settings), Normal-Form Coarse
Correlated Equilibria (NFCCE), and Extensive-Form Correlated Equilibria (EFCE)
in EFGs. We prove that, in those settings, the \emph{$\Phi$-Hedge} algorithms
are equivalent to standard Online Mirror Descent (OMD) algorithms for EFGs with
suitable dilated regularizers, and run in polynomial time. This new connection
further allows us to design and analyze a new class of OMD algorithms based on
modifying its log-partition function. In particular, we design an improved
algorithm with balancing techniques that achieves a sharp
$\widetilde{\mathcal{O}}(\sqrt{XAT})$ EFCE-regret under bandit-feedback in an
EFG with $X$ information sets, $A$ actions, and $T$ episodes. To our best
knowledge, this is the first such rate and matches the information-theoretic
lower bound. |
3,387 | Zero-Shot and Few-Shot Learning for Lung Cancer Multi-Label Classification using Vision Transformer | Lung cancer is the leading cause of cancer-related death worldwide. Lung
adenocarcinoma (LUAD) and lung squamous cell carcinoma (LUSC) are the most
common histologic subtypes of non-small-cell lung cancer (NSCLC). Histology is
an essential tool for lung cancer diagnosis. Pathologists make classifications
according to the dominant subtypes. Although morphology remains the standard
for diagnosis, significant tool needs to be developed to elucidate the
diagnosis. In our study, we utilize the pre-trained Vision Transformer (ViT)
model to classify multiple label lung cancer on histologic slices (from dataset
LC25000), in both Zero-Shot and Few-Shot settings. Then we compare the
performance of Zero-Shot and Few-Shot ViT on accuracy, precision, recall,
sensitivity and specificity. Our study show that the pre-trained ViT model has
a good performance in Zero-Shot setting, a competitive accuracy ($99.87\%$) in
Few-Shot setting ({epoch = 1}) and an optimal result ($100.00\%$ on both
validation set and test set) in Few-Shot seeting ({epoch = 5}). |
3,388 | Self-Supervised Visual Representation Learning with Semantic Grouping | In this paper, we tackle the problem of learning visual representations from
unlabeled scene-centric data. Existing works have demonstrated the potential of
utilizing the underlying complex structure within scene-centric data; still,
they commonly rely on hand-crafted objectness priors or specialized pretext
tasks to build a learning framework, which may harm generalizability. Instead,
we propose contrastive learning from data-driven semantic slots, namely
SlotCon, for joint semantic grouping and representation learning. The semantic
grouping is performed by assigning pixels to a set of learnable prototypes,
which can adapt to each sample by attentive pooling over the feature and form
new slots. Based on the learned data-dependent slots, a contrastive objective
is employed for representation learning, which enhances the discriminability of
features, and conversely facilitates grouping semantically coherent pixels
together. Compared with previous efforts, by simultaneously optimizing the two
coupled objectives of semantic grouping and contrastive learning, our approach
bypasses the disadvantages of hand-crafted priors and is able to learn
object/group-level representations from scene-centric images. Experiments show
our approach effectively decomposes complex scenes into semantic groups for
feature learning and significantly benefits downstream tasks, including object
detection, instance segmentation, and semantic segmentation. The code will be
made publicly available. |
3,389 | Testing for Geometric Invariance and Equivariance | Invariant and equivariant models incorporate the symmetry of an object to be
estimated (here non-parametric regression functions $f : \mathcal{X}
\rightarrow \mathbb{R}$). These models perform better (with respect to $L^2$
loss) and are increasingly being used in practice, but encounter problems when
the symmetry is falsely assumed. In this paper we present a framework for
testing for $G$-equivariance for any semi-group $G$. This will give confidence
to the use of such models when the symmetry is not known a priori. These tests
are independent of the model and are computationally quick, so can be easily
used before model fitting to test their validity. |
3,390 | MetaSSD: Meta-Learned Self-Supervised Detection | Deep learning-based symbol detector gains increasing attention due to the
simple algorithm design than the traditional model-based algorithms such as
Viterbi and BCJR. The supervised learning framework is often employed to
predict the input symbols, where training symbols are used to train the model.
There are two major limitations in the supervised approaches: a) a model needs
to be retrained from scratch when new train symbols come to adapt to a new
channel status, and b) the length of the training symbols needs to be longer
than a certain threshold to make the model generalize well on unseen symbols.
To overcome these challenges, we propose a meta-learning-based self-supervised
symbol detector named MetaSSD. Our contribution is two-fold: a) meta-learning
helps the model adapt to a new channel environment based on experience with
various meta-training environments, and b) self-supervised learning helps the
model to use relatively less supervision than the previously suggested
learning-based detectors. In experiments, MetaSSD outperforms OFDM-MMSE with
noisy channel information and shows comparable results with BCJR. Further
ablation studies show the necessity of each component in our framework. |
3,391 | Kernel Neural Optimal Transport | We study the Neural Optimal Transport (NOT) algorithm which uses the general
optimal transport formulation and learns stochastic transport plans. We show
that NOT with the weak quadratic cost might learn fake plans which are not
optimal. To resolve this issue, we introduce kernel weak quadratic costs. We
show that they provide improved theoretical guarantees and practical
performance. We test NOT with kernel costs on the unpaired image-to-image
translation task. |
3,392 | Federated X-Armed Bandit | This work establishes the first framework of federated $\mathcal{X}$-armed
bandit, where different clients face heterogeneous local objective functions
defined on the same domain and are required to collaboratively figure out the
global optimum. We propose the first federated algorithm for such problems,
named \texttt{Fed-PNE}. By utilizing the topological structure of the global
objective inside the hierarchical partitioning and the weak smoothness
property, our algorithm achieves sublinear cumulative regret with respect to
both the number of clients and the evaluation budget. Meanwhile, it only
requires logarithmic communications between the central server and clients,
protecting the client privacy. Experimental results on synthetic functions and
real datasets validate the advantages of \texttt{Fed-PNE} over single-client
algorithms and federated multi-armed bandit algorithms. |
3,393 | Going Beyond One-Hot Encoding in Classification: Can Human Uncertainty Improve Model Performance? | Technological and computational advances continuously drive forward the broad
field of deep learning. In recent years, the derivation of quantities
describing theuncertainty in the prediction - which naturally accompanies the
modeling process - has sparked general interest in the deep learning community.
Often neglected in the machine learning setting is the human uncertainty that
influences numerous labeling processes. As the core of this work, label
uncertainty is explicitly embedded into the training process via distributional
labels. We demonstrate the effectiveness of our approach on image
classification with a remote sensing data set that contains multiple label
votes by domain experts for each image: The incorporation of label uncertainty
helps the model to generalize better to unseen data and increases model
performance. Similar to existing calibration methods, the distributional labels
lead to better-calibrated probabilities, which in turn yield more certain and
trustworthy predictions. |
3,394 | bsnsing: A decision tree induction method based on recursive optimal boolean rule composition | This paper proposes a new mixed-integer programming (MIP) formulation to
optimize split rule selection in the decision tree induction process, and
develops an efficient search algorithm that is able to solve practical
instances of the MIP model faster than commercial solvers. The formulation is
novel for it directly maximizes the Gini reduction, an effective split
selection criterion which has never been modeled in a mathematical program for
its nonconvexity. The proposed approach differs from other optimal
classification tree models in that it does not attempt to optimize the whole
tree, therefore the flexibility of the recursive partitioning scheme is
retained and the optimization model is more amenable. The approach is
implemented in an open-source R package named bsnsing. Benchmarking experiments
on 75 open data sets suggest that bsnsing trees are the most capable of
discriminating new cases compared to trees trained by other decision tree codes
including the rpart, C50, party and tree packages in R. Compared to other
optimal decision tree packages, including DL8.5, OSDT, GOSDT and indirectly
more, bsnsing stands out in its training speed, ease of use and broader
applicability without losing in prediction accuracy. |
3,395 | Pooling Revisited: Your Receptive Field is Suboptimal | The size and shape of the receptive field determine how the network
aggregates local information and affect the overall performance of a model
considerably. Many components in a neural network, such as kernel sizes and
strides for convolution and pooling operations, influence the configuration of
a receptive field. However, they still rely on hyperparameters, and the
receptive fields of existing models result in suboptimal shapes and sizes.
Hence, we propose a simple yet effective Dynamically Optimized Pooling
operation, referred to as DynOPool, which optimizes the scale factors of
feature maps end-to-end by learning the desirable size and shape of its
receptive field in each layer. Any kind of resizing modules in a deep neural
network can be replaced by the operations with DynOPool at a minimal cost.
Also, DynOPool controls the complexity of a model by introducing an additional
loss term that constrains computational cost. Our experiments show that the
models equipped with the proposed learnable resizing module outperform the
baseline networks on multiple datasets in image classification and semantic
segmentation. |
3,396 | Residual Q-Networks for Value Function Factorizing in Multi-Agent Reinforcement Learning | Multi-Agent Reinforcement Learning (MARL) is useful in many problems that
require the cooperation and coordination of multiple agents. Learning optimal
policies using reinforcement learning in a multi-agent setting can be very
difficult as the number of agents increases. Recent solutions such as Value
Decomposition Networks (VDN), QMIX, QTRAN and QPLEX adhere to the centralized
training and decentralized execution scheme and perform factorization of the
joint action-value functions. However, these methods still suffer from
increased environmental complexity, and at times fail to converge in a stable
manner. We propose a novel concept of Residual Q-Networks (RQNs) for MARL,
which learns to transform the individual Q-value trajectories in a way that
preserves the Individual-Global-Max criteria (IGM), but is more robust in
factorizing action-value functions. The RQN acts as an auxiliary network that
accelerates convergence and will become obsolete as the agents reach the
training objectives. The performance of the proposed method is compared against
several state-of-the-art techniques such as QPLEX, QMIX, QTRAN and VDN, in a
range of multi-agent cooperative tasks. The results illustrate that the
proposed method, in general, converges faster, with increased stability and
shows robust performance in a wider family of environments. The improvements in
results are more prominent in environments with severe punishments for
non-cooperative behaviours and especially in the absence of complete state
information during training time. |
3,397 | Re-parameterizing Your Optimizers rather than Architectures | The well-designed structures in neural networks reflect the prior knowledge
incorporated into the models. However, though different models have various
priors, we are used to training them with model-agnostic optimizers (e.g.,
SGD). In this paper, we propose a novel paradigm of incorporating
model-specific prior knowledge into optimizers and using them to train generic
(simple) models. As an implementation, we propose a novel methodology to add
prior knowledge by modifying the gradients according to a set of model-specific
hyper-parameters, which is referred to as Gradient Re-parameterization, and the
optimizers are named RepOptimizers. For the extreme simplicity of model
structure, we focus on a VGG-style plain model and showcase that such a simple
model trained with a RepOptimizer, which is referred to as RepOpt-VGG, performs
on par with the recent well-designed models. From a practical perspective,
RepOpt-VGG is a favorable base model because of its simple structure, high
inference speed and training efficiency. Compared to Structural
Re-parameterization, which adds priors into models via constructing extra
training-time structures, RepOptimizers require no extra forward/backward
computations and solve the problem of quantization. The code and models are
publicly available at https://github.com/DingXiaoH/RepOptimizers. |
3,398 | Multi-Game Decision Transformers | A longstanding goal of the field of AI is a strategy for compiling diverse
experience into a highly capable, generalist agent. In the subfields of vision
and language, this was largely achieved by scaling up transformer-based models
and training them on large, diverse datasets. Motivated by this progress, we
investigate whether the same strategy can be used to produce generalist
reinforcement learning agents. Specifically, we show that a single
transformer-based model - with a single set of weights - trained purely offline
can play a suite of up to 46 Atari games simultaneously at close-to-human
performance. When trained and evaluated appropriately, we find that the same
trends observed in language and vision hold, including scaling of performance
with model size and rapid adaptation to new games via fine-tuning. We compare
several approaches in this multi-game setting, such as online and offline RL
methods and behavioral cloning, and find that our Multi-Game Decision
Transformer models offer the best scalability and performance. We release the
pre-trained models and code to encourage further research in this direction.
Additional information, videos and code can be seen at:
sites.google.com/view/multi-game-transformers |
3,399 | Conformal Credal Self-Supervised Learning | In semi-supervised learning, the paradigm of self-training refers to the idea
of learning from pseudo-labels suggested by the learner itself. Across various
domains, corresponding methods have proven effective and achieve
state-of-the-art performance. However, pseudo-labels typically stem from ad-hoc
heuristics, relying on the quality of the predictions though without
guaranteeing their validity. One such method, so-called credal self-supervised
learning, maintains pseudo-supervision in the form of sets of (instead of
single) probability distributions over labels, thereby allowing for a flexible
yet uncertainty-aware labeling. Again, however, there is no justification
beyond empirical effectiveness. To address this deficiency, we make use of
conformal prediction, an approach that comes with guarantees on the validity of
set-valued predictions. As a result, the construction of credal sets of labels
is supported by a rigorous theoretical foundation, leading to better calibrated
and less error-prone supervision for unlabeled data. Along with this, we
present effective algorithms for learning from credal self-supervision. An
empirical study demonstrates excellent calibration properties of the
pseudo-supervision, as well as the competitiveness of our method on several
benchmark datasets. |
Subsets and Splits
No saved queries yet
Save your SQL queries to embed, download, and access them later. Queries will appear here once saved.