Unnamed: 0
int64 0
5k
| title
stringlengths 9
210
| abstract
stringlengths 164
1.92k
|
|---|---|---|
3,900 | Learning black- and gray-box chemotactic PDEs/closures from agent based Monte Carlo simulation data | We propose a machine learning framework for the data-driven discovery of
macroscopic chemotactic Partial Differential Equations (PDEs) -- and the
closures that lead to them -- from high-fidelity, individual-based stochastic
simulations of E.coli bacterial motility. The fine scale, detailed, hybrid
(continuum - Monte Carlo) simulation model embodies the underlying biophysics,
and its parameters are informed from experimental observations of individual
cells. We exploit Automatic Relevance Determination (ARD) within a Gaussian
Process framework for the identification of a parsimonious set of collective
observables that parametrize the law of the effective PDEs. Using these
observables, in a second step we learn effective, coarse-grained "Keller-Segel
class" chemotactic PDEs using machine learning regressors: (a) (shallow)
feedforward neural networks and (b) Gaussian Processes. The learned laws can be
black-box (when no prior knowledge about the PDE law structure is assumed) or
gray-box when parts of the equation (e.g. the pure diffusion part) is known and
"hardwired" in the regression process. We also discuss data-driven corrections
(both additive and functional) of analytically known, approximate closures. |
3,901 | GraphPMU: Event Clustering via Graph Representation Learning Using Locationally-Scarce Distribution-Level Fundamental and Harmonic PMU Measurements | This paper is concerned with the complex task of identifying the type and
cause of the events that are captured by distribution-level phasor measurement
units (D-PMUs) in order to enhance situational awareness in power distribution
systems. Our goal is to address two fundamental challenges in this field: a)
scarcity in measurement locations due to the high cost of purchasing,
installing, and streaming data from D-PMUs; b) limited prior knowledge about
the event signatures due to the fact that the events are diverse, infrequent,
and inherently unscheduled. To tackle these challenges, we propose an
unsupervised graph-representation learning method, called GraphPMU, to
significantly improve the performance in event clustering under
locationally-scarce data availability by proposing the following two new
directions: 1) using the topological information about the relative location of
the few available phasor measurement units on the graph of the power
distribution network; 2) utilizing not only the commonly used fundamental
phasor measurements, bus also the less explored harmonic phasor measurements in
the process of analyzing the signatures of various events. Through a detailed
analysis of several case studies, we show that GraphPMU can highly outperform
the prevalent methods in the literature. |
3,902 | Contextual Pandora's Box | Pandora's Box is a fundamental stochastic optimization problem, where the
decision-maker must find a good alternative while minimizing the search cost of
exploring the value of each alternative. In the original formulation, it is
assumed that accurate priors are given for the values of all the alternatives,
while recent work studies the online variant of Pandora's Box where priors are
originally unknown. In this work, we extend Pandora's Box to the online
setting, while incorporating context. At every round, we are presented with a
number of alternatives each having a context, an exploration cost and an
unknown value drawn from an unknown prior distribution that may change at every
round. Our main result is a no-regret algorithm that performs comparably well
to the optimal algorithm which knows all prior distributions exactly. Our
algorithm works even in the bandit setting where the algorithm never learns the
values of the alternatives that were not explored. The key technique that
enables our result is novel a modification of the realizability condition in
contextual bandits that connects a context to the reservation value of the
corresponding distribution rather than its mean |
3,903 | Learning to segment with limited annotations: Self-supervised pretraining with regression and contrastive loss in MRI | Obtaining manual annotations for large datasets for supervised training of
deep learning (DL) models is challenging. The availability of large unlabeled
datasets compared to labeled ones motivate the use of self-supervised
pretraining to initialize DL models for subsequent segmentation tasks. In this
work, we consider two pre-training approaches for driving a DL model to learn
different representations using: a) regression loss that exploits spatial
dependencies within an image and b) contrastive loss that exploits semantic
similarity between pairs of images. The effect of pretraining techniques is
evaluated in two downstream segmentation applications using Magnetic Resonance
(MR) images: a) liver segmentation in abdominal T2-weighted MR images and b)
prostate segmentation in T2-weighted MR images of the prostate. We observed
that DL models pretrained using self-supervision can be finetuned for
comparable performance with fewer labeled datasets. Additionally, we also
observed that initializing the DL model using contrastive loss based
pretraining performed better than the regression loss. |
3,904 | Trainable Weight Averaging for Fast Convergence and Better Generalization | Stochastic gradient descent (SGD) and its variants are commonly considered as
the de-facto methods to train deep neural networks (DNNs). While recent
improvements to SGD mainly focus on the descent algorithm itself, few works pay
attention to utilizing the historical solutions -- as an iterative method, SGD
has actually gone through substantial explorations before its final
convergence. Recently, an interesting attempt is stochastic weight averaging
(SWA), which significantly improves the generalization by simply averaging the
solutions at the tail stage of training. In this paper, we propose to optimize
the averaging coefficients, leading to our Trainable Weight Averaging (TWA),
essentially a novel training method in a reduced subspace spanned by historical
solutions. TWA is quite efficient and has good generalization capability as the
degree of freedom for training is small. It largely reduces the estimation
error from SWA, making it not only further improve the SWA solutions but also
take full advantage of the solutions generated in the head of training where
SWA fails. In the extensive numerical experiments, (i) TWA achieves consistent
improvements over SWA with less sensitivity to learning rate; (ii) applying TWA
in the head stage of training largely speeds up the convergence, resulting in
over 40% time saving on CIFAR and 30% on ImageNet with improved generalization
compared with regular training. The code is released at
https://github.com/nblt/TWA. |
3,905 | Deep-XFCT: Deep learning 3D-mineral liberation analysis with micro X-ray fluorescence and computed tomography | The rapid development of X-ray micro-computed tomography (micro-CT) opens new
opportunities for 3D analysis of particle and grain-size characterisation,
determination of particle densities and shape factors, estimation of mineral
associations and liberation and locking. Current practices in mineral
liberation analysis are based on 2D representations leading to systematic
errors in the extrapolation to volumetric properties. New quantitative methods
based on tomographic data are therefore urgently required for characterisation
of mineral deposits, mineral processing, characterisation of tailings, rock
typing, stratigraphic refinement, reservoir characterisation for applications
in the resource industry, environmental and material sciences. To date, no
simple non-destructive method exists for 3D mineral liberation analysis. We
present a new development based on combining micro-CT with micro-X-ray
fluorescence (micro-XRF) using deep learning. We demonstrate successful
semi-automated multi-modal analysis of a crystalline magmatic rock where the
new technique overcomes the difficult task of differentiating feldspar from
quartz in micro-CT data set. The approach is universal and can be extended to
any multi-modal and multi-instrument analysis for further refinement. We
conclude that the combination of micro-CT and micro-XRF already provides a new
opportunity for robust 3D mineral liberation analysis in both field and
laboratory applications. |
3,906 | Optimal Neural Network Approximation of Wasserstein Gradient Direction via Convex Optimization | The computation of Wasserstein gradient direction is essential for posterior
sampling problems and scientific computing. The approximation of the
Wasserstein gradient with finite samples requires solving a variational
problem. We study the variational problem in the family of two-layer networks
with squared-ReLU activations, towards which we derive a semi-definite
programming (SDP) relaxation. This SDP can be viewed as an approximation of the
Wasserstein gradient in a broader function family including two-layer networks.
By solving the convex SDP, we obtain the optimal approximation of the
Wasserstein gradient direction in this class of functions. Numerical
experiments including PDE-constrained Bayesian inference and parameter
estimation in COVID-19 modeling demonstrate the effectiveness of the proposed
method. |
3,907 | Undersampling is a Minimax Optimal Robustness Intervention in Nonparametric Classification | While a broad range of techniques have been proposed to tackle distribution
shift, the simple baseline of training on an $\textit{undersampled}$ dataset
often achieves close to state-of-the-art-accuracy across several popular
benchmarks. This is rather surprising, since undersampling algorithms discard
excess majority group data. To understand this phenomenon, we ask if learning
is fundamentally constrained by a lack of minority group samples. We prove that
this is indeed the case in the setting of nonparametric binary classification.
Our results show that in the worst case, an algorithm cannot outperform
undersampling unless there is a high degree of overlap between the train and
test distributions (which is unlikely to be the case in real-world datasets),
or if the algorithm leverages additional structure about the distribution
shift. In particular, in the case of label shift we show that there is always
an undersampling algorithm that is minimax optimal. While in the case of
group-covariate shift we show that there is an undersampling algorithm that is
minimax optimal when the overlap between the group distributions is small. We
also perform an experimental case study on a label shift dataset and find that
in line with our theory the test accuracy of robust neural network classifiers
is constrained by the number of minority samples. |
3,908 | Identifying Patient-Specific Root Causes with the Heteroscedastic Noise Model | Complex diseases are caused by a multitude of factors that may differ between
patients even within the same diagnostic category. A few underlying root causes
may nevertheless initiate the development of disease within each patient. We
therefore focus on identifying patient-specific root causes of disease, which
we equate to the sample-specific predictivity of the exogenous error terms in a
structural equation model. We generalize from the linear setting to the
heteroscedastic noise model where $Y = m(X) + \varepsilon\sigma(X)$ with
non-linear functions $m(X)$ and $\sigma(X)$ representing the conditional mean
and mean absolute deviation, respectively. This model preserves identifiability
but introduces non-trivial challenges that require a customized algorithm
called Generalized Root Causal Inference (GRCI) to extract the error terms
correctly. GRCI recovers patient-specific root causes more accurately than
existing alternatives. |
3,909 | BRIGHT -- Graph Neural Networks in Real-Time Fraud Detection | Detecting fraudulent transactions is an essential component to control risk
in e-commerce marketplaces. Apart from rule-based and machine learning filters
that are already deployed in production, we want to enable efficient real-time
inference with graph neural networks (GNNs), which is useful to catch multihop
risk propagation in a transaction graph. However, two challenges arise in the
implementation of GNNs in production. First, future information in a dynamic
graph should not be considered in message passing to predict the past. Second,
the latency of graph query and GNN model inference is usually up to hundreds of
milliseconds, which is costly for some critical online services. To tackle
these challenges, we propose a Batch and Real-time Inception GrapH Topology
(BRIGHT) framework to conduct an end-to-end GNN learning that allows efficient
online real-time inference. BRIGHT framework consists of a graph transformation
module (Two-Stage Directed Graph) and a corresponding GNN architecture (Lambda
Neural Network). The Two-Stage Directed Graph guarantees that the information
passed through neighbors is only from the historical payment transactions. It
consists of two subgraphs representing historical relationships and real-time
links, respectively. The Lambda Neural Network decouples inference into two
stages: batch inference of entity embeddings and real-time inference of
transaction prediction. Our experiments show that BRIGHT outperforms the
baseline models by >2\% in average w.r.t.~precision. Furthermore, BRIGHT is
computationally efficient for real-time fraud detection. Regarding end-to-end
performance (including neighbor query and inference), BRIGHT can reduce the P99
latency by >75\%. For the inference stage, our speedup is on average
7.8$\times$ compared to the traditional GNN. |
3,910 | Factorized Structured Regression for Large-Scale Varying Coefficient Models | Recommender Systems (RS) pervade many aspects of our everyday digital life.
Proposed to work at scale, state-of-the-art RS allow the modeling of thousands
of interactions and facilitate highly individualized recommendations.
Conceptually, many RS can be viewed as instances of statistical regression
models that incorporate complex feature effects and potentially non-Gaussian
outcomes. Such structured regression models, including time-aware varying
coefficients models, are, however, limited in their applicability to
categorical effects and inclusion of a large number of interactions. Here, we
propose Factorized Structured Regression (FaStR) for scalable varying
coefficient models. FaStR overcomes limitations of general regression models
for large-scale data by combining structured additive regression and
factorization approaches in a neural network-based model implementation. This
fusion provides a scalable framework for the estimation of statistical models
in previously infeasible data settings. Empirical results confirm that the
estimation of varying coefficients of our approach is on par with
state-of-the-art regression techniques, while scaling notably better and also
being competitive with other time-aware RS in terms of prediction performance.
We illustrate FaStR's performance and interpretability on a large-scale
behavioral study with smartphone user data. |
3,911 | Learning to Query Internet Text for Informing Reinforcement Learning Agents | Generalization to out of distribution tasks in reinforcement learning is a
challenging problem. One successful approach improves generalization by
conditioning policies on task or environment descriptions that provide
information about the current transition or reward functions. Previously, these
descriptions were often expressed as generated or crowd sourced text. In this
work, we begin to tackle the problem of extracting useful information from
natural language found in the wild (e.g. internet forums, documentation, and
wikis). These natural, pre-existing sources are especially challenging, noisy,
and large and present novel challenges compared to previous approaches. We
propose to address these challenges by training reinforcement learning agents
to learn to query these sources as a human would, and we experiment with how
and when an agent should query. To address the \textit{how}, we demonstrate
that pretrained QA models perform well at executing zero-shot queries in our
target domain. Using information retrieved by a QA model, we train an agent to
learn \textit{when} it should execute queries. We show that our method
correctly learns to execute queries to maximize reward in a reinforcement
learning setting. |
3,912 | Entropy Maximization with Depth: A Variational Principle for Random Neural Networks | To understand the essential role of depth in neural networks, we investigate
a variational principle for depth: Does increasing depth perform an implicit
optimization for the representations in neural networks? We prove that random
neural networks equipped with batch normalization maximize the differential
entropy of representations with depth up to constant factors, assuming that the
representations are contractive. Thus, representations inherently obey the
\textit{principle of maximum entropy} at initialization, in the absence of
information about the learning task. Our variational formulation for neural
representations characterizes the interplay between representation entropy and
architectural components, including depth, width, and non-linear activations,
thereby potentially inspiring the design of neural architectures. |
3,913 | Tight Lower Bounds on Worst-Case Guarantees for Zero-Shot Learning with Attributes | We develop a rigorous mathematical analysis of zero-shot learning with
attributes. In this setting, the goal is to label novel classes with no
training data, only detectors for attributes and a description of how those
attributes are correlated with the target classes, called the class-attribute
matrix. We develop the first non-trivial lower bound on the worst-case error of
the best map from attributes to classes for this setting, even with perfect
attribute detectors. The lower bound characterizes the theoretical intrinsic
difficulty of the zero-shot problem based on the available information -- the
class-attribute matrix -- and the bound is practically computable from it. Our
lower bound is tight, as we show that we can always find a randomized map from
attributes to classes whose expected error is upper bounded by the value of the
lower bound. We show that our analysis can be predictive of how standard
zero-shot methods behave in practice, including which classes will likely be
confused with others. |
3,914 | Forecasting Patient Demand at Urgent Care Clinics using Machine Learning | Urgent care clinics and emergency departments around the world periodically
suffer from extended wait times beyond patient expectations due to inadequate
staffing levels. These delays have been linked with adverse clinical outcomes.
Previous research into forecasting demand this domain has mostly used a
collection of statistical techniques, with machine learning approaches only now
beginning to emerge in recent literature. The forecasting problem for this
domain is difficult and has also been complicated by the COVID-19 pandemic
which has introduced an additional complexity to this estimation due to typical
demand patterns being disrupted. This study explores the ability of machine
learning methods to generate accurate patient presentations at two large urgent
care clinics located in Auckland, New Zealand. A number of machine learning
algorithms were explored in order to determine the most effective technique for
this problem domain, with the task of making forecasts of daily patient demand
three months in advance. The study also performed an in-depth analysis into the
model behaviour in respect to the exploration of which features are most
effective at predicting demand and which features are capable of adaptation to
the volatility caused by the COVID-19 pandemic lockdowns. The results showed
that ensemble-based methods delivered the most accurate and consistent
solutions on average, generating improvements in the range of 23%-27% over the
existing in-house methods for estimating the daily demand. |
3,915 | Semi-supervised Drifted Stream Learning with Short Lookback | In many scenarios, 1) data streams are generated in real time; 2) labeled
data are expensive and only limited labels are available in the beginning; 3)
real-world data is not always i.i.d. and data drift over time gradually; 4) the
storage of historical streams is limited and model updating can only be
achieved based on a very short lookback window. This learning setting limits
the applicability and availability of many Machine Learning (ML) algorithms. We
generalize the learning task under such setting as a semi-supervised drifted
stream learning with short lookback problem (SDSL). SDSL imposes two
under-addressed challenges on existing methods in semi-supervised learning,
continuous learning, and domain adaptation: 1) robust pseudo-labeling under
gradual shifts and 2) anti-forgetting adaptation with short lookback. To tackle
these challenges, we propose a principled and generic generation-replay
framework to solve SDSL. The framework is able to accomplish: 1) robust
pseudo-labeling in the generation step; 2) anti-forgetting adaption in the
replay step. To achieve robust pseudo-labeling, we develop a novel pseudo-label
classification model to leverage supervised knowledge of previously labeled
data, unsupervised knowledge of new data, and, structure knowledge of invariant
label semantics. To achieve adaptive anti-forgetting model replay, we propose
to view the anti-forgetting adaptation task as a flat region search problem. We
propose a novel minimax game-based replay objective function to solve the flat
region search problem and develop an effective optimization solver. Finally, we
present extensive experiments to demonstrate our framework can effectively
address the task of anti-forgetting learning in drifted streams with short
lookback. |
3,916 | Urban Rhapsody: Large-scale exploration of urban soundscapes | Noise is one of the primary quality-of-life issues in urban environments. In
addition to annoyance, noise negatively impacts public health and educational
performance. While low-cost sensors can be deployed to monitor ambient noise
levels at high temporal resolutions, the amount of data they produce and the
complexity of these data pose significant analytical challenges. One way to
address these challenges is through machine listening techniques, which are
used to extract features in attempts to classify the source of noise and
understand temporal patterns of a city's noise situation. However, the
overwhelming number of noise sources in the urban environment and the scarcity
of labeled data makes it nearly impossible to create classification models with
large enough vocabularies that capture the true dynamism of urban soundscapes
In this paper, we first identify a set of requirements in the yet unexplored
domain of urban soundscape exploration. To satisfy the requirements and tackle
the identified challenges, we propose Urban Rhapsody, a framework that combines
state-of-the-art audio representation, machine learning, and visual analytics
to allow users to interactively create classification models, understand noise
patterns of a city, and quickly retrieve and label audio excerpts in order to
create a large high-precision annotated database of urban sound recordings. We
demonstrate the tool's utility through case studies performed by domain experts
using data generated over the five-year deployment of a one-of-a-kind sensor
network in New York City. |
3,917 | RENs: Relevance Encoding Networks | The manifold assumption for high-dimensional data assumes that the data is
generated by varying a set of parameters obtained from a low-dimensional latent
space. Deep generative models (DGMs) are widely used to learn data
representations in an unsupervised way. DGMs parameterize the underlying
low-dimensional manifold in the data space using bottleneck architectures such
as variational autoencoders (VAEs). The bottleneck dimension for VAEs is
treated as a hyperparameter that depends on the dataset and is fixed at design
time after extensive tuning. As the intrinsic dimensionality of most real-world
datasets is unknown, often, there is a mismatch between the intrinsic
dimensionality and the latent dimensionality chosen as a hyperparameter. This
mismatch can negatively contribute to the model performance for representation
learning and sample generation tasks. This paper proposes relevance encoding
networks (RENs): a novel probabilistic VAE-based framework that uses the
automatic relevance determination (ARD) prior in the latent space to learn the
data-specific bottleneck dimensionality. The relevance of each latent dimension
is directly learned from the data along with the other model parameters using
stochastic gradient descent and a reparameterization trick adapted to
non-Gaussian priors. We leverage the concept of DeepSets to capture permutation
invariant statistical properties in both data and latent spaces for relevance
determination. The proposed framework is general and flexible and can be used
for the state-of-the-art VAE models that leverage regularizers to impose
specific characteristics in the latent space (e.g., disentanglement). With
extensive experimentation on synthetic and public image datasets, we show that
the proposed model learns the relevant latent bottleneck dimensionality without
compromising the representation and generation quality of the samples. |
3,918 | Designing an Efficient End-to-end Machine Learning Pipeline for Real-time Empty-shelf Detection | On-Shelf Availability (OSA) of products in retail stores is a critical
business criterion in the fast moving consumer goods and retails sector. When a
product is out-of-stock (OOS) and a customer cannot find it on its designed
shelf, this motivates the customer to store-switching or buying nothing, which
causes fall in future sales and demands. Retailers are employing several
approaches to detect empty shelves and ensure high OSA of products; however,
such methods are generally ineffective and infeasible since they are either
manual, expensive or less accurate. Recently machine learning based solutions
have been proposed, but they suffer from high computational cost and low
accuracy problem due to lack of large annotated datasets of on-shelf products.
Here, we present an elegant approach for designing an end-to-end machine
learning (ML) pipeline for real-time empty shelf detection. Considering the
strong dependency between the quality of ML models and the quality of data, we
focus on the importance of proper data collection, cleaning and correct data
annotation before delving into modeling. Since an empty-shelf detection
solution should be computationally-efficient for real-time predictions, we
explore different run-time optimizations to improve the model performance. Our
dataset contains 1000 images, collected and annotated by following well-defined
guidelines. Our low-latency model achieves a mean average F1-score of 68.5%,
and can process up to 67 images/s on Intel Xeon Gold and up to 860 images/s on
an A100 GPU. |
3,919 | Efficient and Near-Optimal Smoothed Online Learning for Generalized Linear Functions | Due to the drastic gap in complexity between sequential and batch statistical
learning, recent work has studied a smoothed sequential learning setting, where
Nature is constrained to select contexts with density bounded by 1/{\sigma}
with respect to a known measure {\mu}. Unfortunately, for some function
classes, there is an exponential gap between the statistically optimal regret
and that which can be achieved efficiently. In this paper, we give a
computationally efficient algorithm that is the first to enjoy the
statistically optimal log(T/{\sigma}) regret for realizable K-wise linear
classification. We extend our results to settings where the true classifier is
linear in an over-parameterized polynomial featurization of the contexts, as
well as to a realizable piecewise-regression setting assuming access to an
appropriate ERM oracle. Somewhat surprisingly, standard disagreement-based
analyses are insufficient to achieve regret logarithmic in 1/{\sigma}. Instead,
we develop a novel characterization of the geometry of the disagreement region
induced by generalized linear classifiers. Along the way, we develop numerous
technical tools of independent interest, including a general anti-concentration
bound for the determinant of certain matrix averages. |
3,920 | Scalable and Low-Latency Federated Learning with Cooperative Mobile Edge Networking | Federated learning (FL) enables collaborative model training without
centralizing data. However, the traditional FL framework is cloud-based and
suffers from high communication latency. On the other hand, the edge-based FL
framework that relies on an edge server co-located with access point for model
aggregation has low communication latency but suffers from degraded model
accuracy due to the limited coverage of edge server. In light of high-accuracy
but high-latency cloud-based FL and low-latency but low-accuracy edge-based FL,
this paper proposes a new FL framework based on cooperative mobile edge
networking called cooperative federated edge learning (CFEL) to enable both
high-accuracy and low-latency distributed intelligence at mobile edge networks.
Considering the unique two-tier network architecture of CFEL, a novel federated
optimization method dubbed cooperative edge-based federated averaging
(CE-FedAvg) is further developed, wherein each edge server both coordinates
collaborative model training among the devices within its own coverage and
cooperates with other edge servers to learn a shared global model through
decentralized consensus. Experimental results based on benchmark datasets show
that CFEL can largely speed up the convergence speed and reduce the training
time to achieve a target model accuracy compared with prior FL frameworks. |
3,921 | Online Deep Equilibrium Learning for Regularization by Denoising | Plug-and-Play Priors (PnP) and Regularization by Denoising (RED) are
widely-used frameworks for solving imaging inverse problems by computing
fixed-points of operators combining physical measurement models and learned
image priors. While traditional PnP/RED formulations have focused on priors
specified using image denoisers, there is a growing interest in learning
PnP/RED priors that are end-to-end optimal. The recent Deep Equilibrium Models
(DEQ) framework has enabled memory-efficient end-to-end learning of PnP/RED
priors by implicitly differentiating through the fixed-point equations without
storing intermediate activation values. However, the dependence of the
computational/memory complexity of the measurement models in PnP/RED on the
total number of measurements leaves DEQ impractical for many imaging
applications. We propose ODER as a new strategy for improving the efficiency of
DEQ through stochastic approximations of the measurement models. We
theoretically analyze ODER giving insights into its convergence and ability to
approximate the traditional DEQ approach. Our numerical results suggest the
potential improvements in training/testing complexity due to ODER on three
distinct imaging applications. |
3,922 | Near-Optimal Goal-Oriented Reinforcement Learning in Non-Stationary Environments | We initiate the study of dynamic regret minimization for goal-oriented
reinforcement learning modeled by a non-stationary stochastic shortest path
problem with changing cost and transition functions. We start by establishing a
lower bound $\Omega((B_{\star} SAT_{\star}(\Delta_c +
B_{\star}^2\Delta_P))^{1/3}K^{2/3})$, where $B_{\star}$ is the maximum expected
cost of the optimal policy of any episode starting from any state, $T_{\star}$
is the maximum hitting time of the optimal policy of any episode starting from
the initial state, $SA$ is the number of state-action pairs, $\Delta_c$ and
$\Delta_P$ are the amount of changes of the cost and transition functions
respectively, and $K$ is the number of episodes. The different roles of
$\Delta_c$ and $\Delta_P$ in this lower bound inspire us to design algorithms
that estimate costs and transitions separately. Specifically, assuming the
knowledge of $\Delta_c$ and $\Delta_P$, we develop a simple but sub-optimal
algorithm and another more involved minimax optimal algorithm (up to
logarithmic terms). These algorithms combine the ideas of finite-horizon
approximation [Chen et al., 2022a], special Bernstein-style bonuses of the MVP
algorithm [Zhang et al., 2020], adaptive confidence widening [Wei and Luo,
2021], as well as some new techniques such as properly penalizing long-horizon
policies. Finally, when $\Delta_c$ and $\Delta_P$ are unknown, we develop a
variant of the MASTER algorithm [Wei and Luo, 2021] and integrate the
aforementioned ideas into it to achieve $\widetilde{O}(\min\{B_{\star}
S\sqrt{ALK},
(B_{\star}^2S^2AT_{\star}(\Delta_c+B_{\star}\Delta_P))^{1/3}K^{2/3}\})$ regret,
where $L$ is the unknown number of changes of the environment. |
3,923 | Improving Subgraph Representation Learning via Multi-View Augmentation | Subgraph representation learning based on Graph Neural Network (GNN) has
broad applications in chemistry and biology, such as molecule property
prediction and gene collaborative function prediction. On the other hand, graph
augmentation techniques have shown promising results in improving graph-based
and node-based classification tasks but are rarely explored in the GNN-based
subgraph representation learning literature. In this work, we developed a novel
multiview augmentation mechanism to improve subgraph representation learning
and thus the accuracy of downstream prediction tasks. The augmentation
technique creates multiple variants of subgraphs and embeds these variants into
the original graph to achieve both high training efficiency, scalability, and
improved accuracy. Experiments on several real-world subgraph benchmarks
demonstrate the superiority of our proposed multi-view augmentation techniques. |
3,924 | EvoVGM: a Deep Variational Generative Model for Evolutionary Parameter Estimation | Most evolutionary-oriented deep generative models do not explicitly consider
the underlying evolutionary dynamics of biological sequences as it is performed
within the Bayesian phylogenetic inference framework. In this study, we propose
a method for a deep variational Bayesian generative model (EvoVGM) that jointly
approximates the true posterior of local evolutionary parameters and generates
sequence alignments. Moreover, it is instantiated and tuned for continuous-time
Markov chain substitution models such as JC69, K80 and GTR. We train the model
via a low-variance stochastic estimator and a gradient ascent algorithm. Here,
we analyze the consistency and effectiveness of EvoVGM on synthetic sequence
alignments simulated with several evolutionary scenarios and different sizes.
Finally, we highlight the robustness of a fine-tuned EvoVGM model using a
sequence alignment of gene S of coronaviruses. |
3,925 | Topological Simplification of Signals for Inference and Approximate Reconstruction | As Internet of Things (IoT) devices become both cheaper and more powerful,
researchers are increasingly finding solutions to their scientific curiosities
both financially and computationally feasible. When operating with restricted
power or communications budgets, however, devices can only send
highly-compressed data. Such circumstances are common for devices placed away
from electric grids that can only communicate via satellite, a situation
particularly plausible for environmental sensor networks. These restrictions
can be further complicated by potential variability in the communications
budget, for example a solar-powered device needing to expend less energy when
transmitting data on a cloudy day. We propose a novel, topology-based, lossy
compression method well-equipped for these restrictive yet variable
circumstances. This technique, Topological Signal Compression, allows sending
compressed signals that utilize the entirety of a variable communications
budget. To demonstrate our algorithm's capabilities, we perform entropy
calculations as well as a classification exercise on increasingly topologically
simplified signals from the Free-Spoken Digit Dataset and explore the stability
of the resulting performance against common baselines. |
3,926 | Concurrent Neural Tree and Data Preprocessing AutoML for Image Classification | Deep Neural Networks (DNN's) are a widely-used solution for a variety of
machine learning problems. However, it is often necessary to invest a
significant amount of a data scientist's time to pre-process input data, test
different neural network architectures, and tune hyper-parameters for optimal
performance. Automated machine learning (autoML) methods automatically search
the architecture and hyper-parameter space for optimal neural networks.
However, current state-of-the-art (SOTA) methods do not include traditional
methods for manipulating input data as part of the algorithmic search space. We
adapt the Evolutionary Multi-objective Algorithm Design Engine (EMADE), a
multi-objective evolutionary search framework for traditional machine learning
methods, to perform neural architecture search. We also integrate EMADE's
signal processing and image processing primitives. These primitives allow EMADE
to manipulate input data before ingestion into the simultaneously evolved DNN.
We show that including these methods as part of the search space shows
potential to provide benefits to performance on the CIFAR-10 image
classification benchmark dataset. |
3,927 | Formalizing Preferences Over Runtime Distributions | When trying to solve a computational problem we are often faced with a choice
among algorithms that are all guaranteed to return the right answer but that
differ in their runtime distributions (e.g., SAT solvers, sorting algorithms).
This paper aims to lay theoretical foundations for such choices by formalizing
preferences over runtime distributions. It might seem that we should simply
prefer the algorithm that minimizes expected runtime. However, such preferences
would be driven by exactly how slow our algorithm is on bad inputs, whereas in
practice we are typically willing to cut off occasional, sufficiently long runs
before they finish. We propose a principled alternative, taking a
utility-theoretic approach to characterize the scoring functions that describe
preferences over algorithms. These functions depend on the way our value for
solving our problem decreases with time and on the distribution from which
captimes are drawn. We describe examples of realistic utility functions and
show how to leverage a maximum-entropy approach for modeling underspecified
captime distributions. Finally, we show how to efficiently estimate an
algorithm's expected utility from runtime samples. |
3,928 | Preference Dynamics Under Personalized Recommendations | Many projects (both practical and academic) have designed algorithms to match
users to content they will enjoy under the assumption that user's preferences
and opinions do not change with the content they see. Evidence suggests that
individuals' preferences are directly shaped by what content they see --
radicalization, rabbit holes, polarization, and boredom are all example
phenomena of preferences affected by content. Polarization in particular can
occur even in ecosystems with "mass media," where no personalization takes
place, as recently explored in a natural model of preference dynamics
by~\citet{hkazla2019geometric} and~\citet{gaitonde2021polarization}. If all
users' preferences are drawn towards content they already like, or are repelled
from content they already dislike, uniform consumption of media leads to a
population of heterogeneous preferences converging towards only two poles.
In this work, we explore whether some phenomenon akin to polarization occurs
when users receive \emph{personalized} content recommendations. We use a
similar model of preference dynamics, where an individual's preferences move
towards content the consume and enjoy, and away from content they consume and
dislike. We show that standard user reward maximization is an almost trivial
goal in such an environment (a large class of simple algorithms will achieve
only constant regret). A more interesting objective, then, is to understand
under what conditions a recommendation algorithm can ensure stationarity of
user's preferences. We show how to design a content recommendations which can
achieve approximate stationarity, under mild conditions on the set of available
content, when a user's preferences are known, and how one can learn enough
about a user's preferences to implement such a strategy even when user
preferences are initially unknown. |
3,929 | People counting system for retail analytics using edge AI | Developments in IoT applications are playing an important role in our
day-to-day life, starting from business predictions to self driving cars. One
of the area, most influenced by the field of AI and IoT is retail analytics. In
Retail Analytics, Conversion Rates - a metric which is most often used by
retail stores to measure how many people have visited the store and how many
purchases has happened. This retail conversion rate assess the marketing
operations, increasing stock, store outlet and running promotions ..etc. Our
project intends to build a cost-effective people counting system with AI at
Edge, where it calculates Conversion rates using total number of people counted
by the system and number of transactions for the day, which helps in providing
analytical insights for retail store optimization with a very minimum hardware
requirements. |
3,930 | BiT: Robustly Binarized Multi-distilled Transformer | Modern pre-trained transformers have rapidly advanced the state-of-the-art in
machine learning, but have also grown in parameters and computational
complexity, making them increasingly difficult to deploy in
resource-constrained environments. Binarization of the weights and activations
of the network can significantly alleviate these issues, however is technically
challenging from an optimization perspective. In this work, we identify a
series of improvements which enables binary transformers at a much higher
accuracy than what was possible previously. These include a two-set
binarization scheme, a novel elastic binary activation function with learned
parameters, and a method to quantize a network to its limit by successively
distilling higher precision models into lower precision students. These
approaches allow for the first time, fully binarized transformer models that
are at a practical level of accuracy, approaching a full-precision BERT
baseline on the GLUE language understanding benchmark within as little as 5.9%. |
3,931 | TSEM: Temporally Weighted Spatiotemporal Explainable Neural Network for Multivariate Time Series | Deep learning has become a one-size-fits-all solution for technical and
business domains thanks to its flexibility and adaptability. It is implemented
using opaque models, which unfortunately undermines the outcome
trustworthiness. In order to have a better understanding of the behavior of a
system, particularly one driven by time series, a look inside a deep learning
model so-called posthoc eXplainable Artificial Intelligence (XAI) approaches,
is important. There are two major types of XAI for time series data, namely
model-agnostic and model-specific. Model-specific approach is considered in
this work. While other approaches employ either Class Activation Mapping (CAM)
or Attention Mechanism, we merge the two strategies into a single system,
simply called the Temporally Weighted Spatiotemporal Explainable Neural Network
for Multivariate Time Series (TSEM). TSEM combines the capabilities of RNN and
CNN models in such a way that RNN hidden units are employed as attention
weights for the CNN feature maps temporal axis. The result shows that TSEM
outperforms XCM. It is similar to STAM in terms of accuracy, while also
satisfying a number of interpretability criteria, including causality,
fidelity, and spatiotemporality. |
3,932 | QGNN: Value Function Factorisation with Graph Neural Networks | In multi-agent reinforcement learning, the use of a global objective is a
powerful tool for incentivising cooperation. Unfortunately, it is not
sample-efficient to train individual agents with a global reward, because it
does not necessarily correlate with an agent's individual actions. This problem
can be solved by factorising the global value function into local value
functions. Early work in this domain performed factorisation by conditioning
local value functions purely on local information. Recently, it has been shown
that providing both local information and an encoding of the global state can
promote cooperative behaviour. In this paper we propose QGNN, the first value
factorisation method to use a graph neural network (GNN) based model. The
multi-layer message passing architecture of QGNN provides more representational
complexity than models in prior work, allowing it to produce a more effective
factorisation. QGNN also introduces a permutation invariant mixer which is able
to match the performance of other methods, even with significantly fewer
parameters. We evaluate our method against several baselines, including
QMIX-Att, GraphMIX, QMIX, VDN, and hybrid architectures. Our experiments
include Starcraft, the standard benchmark for credit assignment; Estimate Game,
a custom environment that explicitly models inter-agent dependencies; and
Coalition Structure Generation, a foundational problem with real-world
applications. The results show that QGNN outperforms state-of-the-art value
factorisation baselines consistently. |
3,933 | Inception Transformer | Recent studies show that Transformer has strong capability of building
long-range dependencies, yet is incompetent in capturing high frequencies that
predominantly convey local information. To tackle this issue, we present a
novel and general-purpose Inception Transformer, or iFormer for short, that
effectively learns comprehensive features with both high- and low-frequency
information in visual data. Specifically, we design an Inception mixer to
explicitly graft the advantages of convolution and max-pooling for capturing
the high-frequency information to Transformers. Different from recent hybrid
frameworks, the Inception mixer brings greater efficiency through a channel
splitting mechanism to adopt parallel convolution/max-pooling path and
self-attention path as high- and low-frequency mixers, while having the
flexibility to model discriminative information scattered within a wide
frequency range. Considering that bottom layers play more roles in capturing
high-frequency details while top layers more in modeling low-frequency global
information, we further introduce a frequency ramp structure, i.e. gradually
decreasing the dimensions fed to the high-frequency mixer and increasing those
to the low-frequency mixer, which can effectively trade-off high- and
low-frequency components across different layers. We benchmark the iFormer on a
series of vision tasks, and showcase that it achieves impressive performance on
image classification, COCO detection and ADE20K segmentation. For example, our
iFormer-S hits the top-1 accuracy of 83.4% on ImageNet-1K, much higher than
DeiT-S by 3.6%, and even slightly better than much bigger model Swin-B (83.3%)
with only 1/4 parameters and 1/3 FLOPs. Code and models will be released at
https://github.com/sail-sg/iFormer. |
3,934 | Federated Adaptation of Reservoirs via Intrinsic Plasticity | We propose a novel algorithm for performing federated learning with Echo
State Networks (ESNs) in a client-server scenario. In particular, our proposal
focuses on the adaptation of reservoirs by combining Intrinsic Plasticity with
Federated Averaging. The former is a gradient-based method for adapting the
reservoir's non-linearity in a local and unsupervised manner, while the latter
provides the framework for learning in the federated scenario. We evaluate our
approach on real-world datasets from human monitoring, in comparison with the
previous approach for federated ESNs existing in literature. Results show that
adapting the reservoir with our algorithm provides a significant improvement on
the performance of the global model. |
3,935 | Learning Mean Field Games: A Survey | Non-cooperative and cooperative games with a very large number of players
have many applications but remain generally intractable when the number of
players increases. Introduced by Lasry and Lions, and Huang, Caines and
Malham\'e, Mean Field Games (MFGs) rely on a mean-field approximation to allow
the number of players to grow to infinity. Traditional methods for solving
these games generally rely on solving partial or stochastic differential
equations with a full knowledge of the model. Recently, Reinforcement Learning
(RL) has appeared promising to solve complex problems. By combining MFGs and
RL, we hope to solve games at a very large scale both in terms of population
size and environment complexity. In this survey, we review the quickly growing
recent literature on RL methods to learn Nash equilibria in MFGs. We first
identify the most common settings (static, stationary, and evolutive). We then
present a general framework for classical iterative methods (based on
best-response computation or policy evaluation) to solve MFGs in an exact way.
Building on these algorithms and the connection with Markov Decision Processes,
we explain how RL can be used to learn MFG solutions in a model-free way. Last,
we present numerical illustrations on a benchmark problem, and conclude with
some perspectives. |
3,936 | Conformal Prediction Intervals with Temporal Dependence | Cross-sectional prediction is common in many domains such as healthcare,
including forecasting tasks using electronic health records, where different
patients form a cross-section. We focus on the task of constructing valid
prediction intervals (PIs) in time-series regression with a cross-section. A
prediction interval is considered valid if it covers the true response with (a
pre-specified) high probability. We first distinguish between two notions of
validity in such a setting: cross-sectional and longitudinal. Cross-sectional
validity is concerned with validity across the cross-section of the time series
data, while longitudinal validity accounts for the temporal dimension. Coverage
guarantees along both these dimensions are ideally desirable; however, we show
that distribution-free longitudinal validity is theoretically impossible.
Despite this limitation, we propose Conformal Prediction with Temporal
Dependence (CPTD), a procedure which is able to maintain strict cross-sectional
validity while improving longitudinal coverage. CPTD is post-hoc and
light-weight, and can easily be used in conjunction with any prediction model
as long as a calibration set is available. We focus on neural networks due to
their ability to model complicated data such as diagnosis codes for time-series
regression, and perform extensive experimental validation to verify the
efficacy of our approach. We find that CPTD outperforms baselines on a variety
of datasets by improving longitudinal coverage and often providing more
efficient (narrower) PIs. |
3,937 | Predicting Corporate Risk by Jointly Modeling Company Networks and Dialogues in Earnings Conference Calls | Earnings conference calls are attracting an increasing number of researchers
due to their free form and rich information. Existing studies, however, do not
take speaker role information into account. Furthermore, current research does
not fully account for the impact of inter-company relationships on company
risk. The only study that integrates company networks and earnings conference
calls constructs an undirected graph for companies holding earnings conference
calls at different dates, failing to meet the requirement of no temporal
information leakage for prediction tasks. To address the aforementioned issues,
we propose a new model called Temporal Virtual Graph Neural Network (TVGNN),
which incorporates earnings conference calls and company networks to predict
company risk. For the first time, our model incorporates participant role
information in dialogue modeling. Moreover, we develop a new approach to
construct company networks that ensures no temporal information leakage in the
graph. In experiments, our proposed model outperforms all baselines. The
supplementary analyses demonstrate the model's effectiveness and
interpretability. |
3,938 | Mitigating multiple descents: A model-agnostic framework for risk monotonization | Recent empirical and theoretical analyses of several commonly used prediction
procedures reveal a peculiar risk behavior in high dimensions, referred to as
double/multiple descent, in which the asymptotic risk is a non-monotonic
function of the limiting aspect ratio of the number of features or parameters
to the sample size. To mitigate this undesirable behavior, we develop a general
framework for risk monotonization based on cross-validation that takes as input
a generic prediction procedure and returns a modified procedure whose
out-of-sample prediction risk is, asymptotically, monotonic in the limiting
aspect ratio. As part of our framework, we propose two data-driven
methodologies, namely zero- and one-step, that are akin to bagging and
boosting, respectively, and show that, under very mild assumptions, they
provably achieve monotonic asymptotic risk behavior. Our results are applicable
to a broad variety of prediction procedures and loss functions, and do not
require a well-specified (parametric) model. We exemplify our framework with
concrete analyses of the minimum $\ell_2$, $\ell_1$-norm least squares
prediction procedures. As one of the ingredients in our analysis, we also
derive novel additive and multiplicative forms of oracle risk inequalities for
split cross-validation that are of independent interest. |
3,939 | Amortized Inference for Causal Structure Learning | Learning causal structure poses a combinatorial search problem that typically
involves evaluating structures using a score or independence test. The
resulting search is costly, and designing suitable scores or tests that capture
prior knowledge is difficult. In this work, we propose to amortize the process
of causal structure learning. Rather than searching over causal structures
directly, we train a variational inference model to predict the causal
structure from observational/interventional data. Our inference model acquires
domain-specific inductive bias for causal discovery solely from data generated
by a simulator. This allows us to bypass both the search over graphs and the
hand-engineering of suitable score functions. Moreover, the architecture of our
inference model is permutation invariant w.r.t. the data points and permutation
equivariant w.r.t. the variables, facilitating generalization to significantly
larger problem instances than seen during training. On synthetic data and
semi-synthetic gene expression data, our models exhibit robust generalization
capabilities under substantial distribution shift and significantly outperform
existing algorithms, especially in the challenging genomics domain. |
3,940 | Analytics of Business Time Series Using Machine Learning and Bayesian Inference | In the survey we consider the case studies on sales time series forecasting,
the deep learning approach for forecasting non-stationary time series using
time trend correction, dynamic price and supply optimization using Q-learning,
Bitcoin price modeling, COVID-19 spread impact on stock market, using social
networks signals in analytics. The use of machine learning and Bayesian
inference in predictive analytics has been analyzed. |
3,941 | A Neural Tangent Kernel Formula for Ensembles of Soft Trees with Arbitrary Architectures | A soft tree is an actively studied variant of a decision tree that updates
splitting rules using the gradient method. Although it can have various tree
architectures, the theoretical properties of their impact are not well known.
In this paper, we formulate and analyze the Neural Tangent Kernel (NTK) induced
by soft tree ensembles for arbitrary tree architectures. This kernel leads to
the remarkable finding that only the number of leaves at each depth is relevant
for the tree architecture in ensemble learning with infinitely many trees. In
other words, if the number of leaves at each depth is fixed, the training
behavior in function space and the generalization performance are exactly the
same across different tree architectures, even if they are not isomorphic. We
also show that the NTK of asymmetric trees like decision lists does not
degenerate when they get infinitely deep. This is in contrast to the perfect
binary trees, whose NTK is known to degenerate and leads to worse
generalization performance for deeper trees. |
3,942 | RADNet: Ensemble Model for Robust Glaucoma Classification in Color Fundus Images | Glaucoma is one of the most severe eye diseases, characterized by rapid
progression and leading to irreversible blindness. It is often the case that
pathology diagnostics is carried out when the one's sight has already
significantly degraded due to the lack of noticeable symptoms at early stage of
the disease. Regular glaucoma screenings of the population shall improve
early-stage detection, however the desirable frequency of etymological checkups
is often not feasible due to excessive load imposed by manual diagnostics on
limited number of specialists. Considering the basic methodology to detect
glaucoma is to analyze fundus images for the \textit{optic-disc-to-optic-cup
ratio}, Machine Learning domain can offer sophisticated tooling for image
processing and classification. In our work, we propose an advanced image
pre-processing technique combined with an ensemble of deep classification
networks. Our \textit{Retinal Auto Detection (RADNet)} model has been
successfully tested on Rotterdam EyePACS AIROGS train dataset with AUC of 0.92,
and then additionally finetuned and tested on a fraction of RIM-ONE DL dataset
with AUC of 0.91. |
3,943 | Differentially Private Data Generation Needs Better Features | Training even moderately-sized generative models with differentially-private
stochastic gradient descent (DP-SGD) is difficult: the required level of noise
for reasonable levels of privacy is simply too large. We advocate instead
building off a good, relevant representation on public data, then using private
data only for "transfer learning." In particular, we minimize the maximum mean
discrepancy (MMD) between private target data and the generated distribution,
using a kernel based on perceptual features from a public dataset. With the
MMD, we can simply privatize the data-dependent term once and for all, rather
than introducing noise at each step of optimization as in DP-SGD. Our algorithm
allows us to generate CIFAR10-level images faithfully with $\varepsilon \approx
2$, far surpassing the current state of the art, which only models MNIST and
FashionMNIST at $\varepsilon \approx 10$. Our work introduces simple yet
powerful foundations for reducing the gap between private and non-private deep
generative models. |
3,944 | Robust Reinforcement Learning on Graphs for Logistics optimization | Logistics optimization nowadays is becoming one of the hottest areas in the
AI community. In the past year, significant advancements in the domain were
achieved by representing the problem in a form of graph. Another promising area
of research was to apply reinforcement learning algorithms to the above task.
In our work, we made advantage of using both approaches and apply reinforcement
learning on a graph. To do that, we have analyzed the most recent results in
both fields and selected SOTA algorithms both from graph neural networks and
reinforcement learning. Then, we combined selected models on the problem of
AMOD systems optimization for the transportation network of New York city. Our
team compared three algorithms - GAT, Pro-CNN and PTDNet - to bring to the fore
the important nodes on a graph representation. Finally, we achieved SOTA
results on AMOD systems optimization problem employing PTDNet with GNN and
training them in reinforcement fashion.
Keywords: Graph Neural Network (GNN), Logistics optimization, Reinforcement
Learning |
3,945 | Trust-based Consensus in Multi-Agent Reinforcement Learning Systems | An often neglected issue in multi-agent reinforcement learning (MARL) is the
potential presence of unreliable agents in the environment whose deviations
from expected behavior can prevent a system from accomplishing its intended
tasks. In particular, consensus is a fundamental underpinning problem of
cooperative distributed multi-agent systems. Consensus requires different
agents, situated in a decentralized communication network, to reach an
agreement out of a set of initial proposals that they put forward.
Learning-based agents should adopt a protocol that allows them to reach
consensus despite having one or more unreliable agents in the system. This
paper investigates the problem of unreliable agents in MARL, considering
consensus as case study. Echoing established results in the distributed systems
literature, our experiments show that even a moderate fraction of such agents
can greatly impact the ability of reaching consensus in a networked
environment. We propose Reinforcement Learning-based Trusted Consensus (RLTC),
a decentralized trust mechanism, in which agents can independently decide which
neighbors to communicate with. We empirically demonstrate that our trust
mechanism is able to deal with unreliable agents effectively, as evidenced by
higher consensus success rates. |
3,946 | Understanding Programmatic Weak Supervision via Source-aware Influence Function | Programmatic Weak Supervision (PWS) aggregates the source votes of multiple
weak supervision sources into probabilistic training labels, which are in turn
used to train an end model. With its increasing popularity, it is critical to
have some tool for users to understand the influence of each component (e.g.,
the source vote or training data) in the pipeline and interpret the end model
behavior. To achieve this, we build on Influence Function (IF) and propose
source-aware IF, which leverages the generation process of the probabilistic
labels to decompose the end model's training objective and then calculate the
influence associated with each (data, source, class) tuple. These primitive
influence score can then be used to estimate the influence of individual
component of PWS, such as source vote, supervision source, and training data.
On datasets of diverse domains, we demonstrate multiple use cases: (1)
interpreting incorrect predictions from multiple angles that reveals insights
for debugging the PWS pipeline, (2) identifying mislabeling of sources with a
gain of 9%-37% over baselines, and (3) improving the end model's generalization
performance by removing harmful components in the training objective (13%-24%
better than ordinary IF). |
3,947 | Image Colorization using U-Net with Skip Connections and Fusion Layer on Landscape Images | We present a novel technique to automatically colorize grayscale images that
combine the U-Net model and Fusion Layer features. This approach allows the
model to learn the colorization of images from pre-trained U-Net. Moreover, the
Fusion layer is applied to merge local information results dependent on small
image patches with global priors of an entire image on each class, forming
visually more compelling colorization results. Finally, we validate our
approach with a user study evaluation and compare it against state-of-the-art,
resulting in improvements. |
3,948 | Stochastic Second-Order Methods Provably Beat SGD For Gradient-Dominated Functions | We study the performance of Stochastic Cubic Regularized Newton (SCRN) on a
class of functions satisfying gradient dominance property which holds in a wide
range of applications in machine learning and signal processing. This condition
ensures that any first-order stationary point is a global optimum. We prove
that SCRN improves the best-known sample complexity of stochastic gradient
descent in achieving $\epsilon$-global optimum by a factor of
$\mathcal{O}(\epsilon^{-1/2})$. Even under a weak version of gradient dominance
property, which is applicable to policy-based reinforcement learning (RL), SCRN
achieves the same improvement over stochastic policy gradient methods.
Additionally, we show that the sample complexity of SCRN can be improved by a
factor of ${\mathcal{O}}(\epsilon^{-1/2})$ using a variance reduction method
with time-varying batch sizes. Experimental results in various RL settings
showcase the remarkable performance of SCRN compared to first-order methods. |
3,949 | A Universal Error Measure for Input Predictions Applied to Online Graph Problems | We introduce a novel measure for quantifying the error in input predictions.
The error is based on a minimum-cost hyperedge cover in a suitably defined
hypergraph and provides a general template which we apply to online graph
problems. The measure captures errors due to absent predicted requests as well
as unpredicted actual requests; hence, predicted and actual inputs can be of
arbitrary size. We achieve refined performance guarantees for previously
studied network design problems in the online-list model, such as Steiner tree
and facility location. Further, we initiate the study of learning-augmented
algorithms for online routing problems, such as the traveling salesperson
problem and dial-a-ride problem, where (transportation) requests arrive over
time (online-time model). We provide a general algorithmic framework and we
give error-dependent performance bounds that improve upon known worst-case
barriers, when given accurate predictions, at the cost of slightly increased
worst-case bounds when given predictions of arbitrary quality. |
3,950 | Removing the fat from your posterior samples with margarine | Bayesian workflows often require the introduction of nuisance parameters, yet
for core science modelling one needs access to a marginal posterior density. In
this work we use masked autoregressive flows and kernel density estimators to
encapsulate the marginal posterior, allowing us to compute marginal
Kullback-Leibler divergences and marginal Bayesian model dimensionalities in
addition to generating samples and computing marginal log probabilities. We
demonstrate this in application to topical cosmological examples of the Dark
Energy Survey, and global 21cm signal experiments. In addition to the
computation of marginal Bayesian statistics, this work is important for further
applications in Bayesian experimental design, complex prior modelling and
likelihood emulation. This technique is made publicly available in the
pip-installable code margarine. |
3,951 | Mirror Descent Maximizes Generalized Margin and Can Be Implemented Efficiently | Driven by the empirical success and wide use of deep neural networks,
understanding the generalization performance of overparameterized models has
become an increasingly popular question. To this end, there has been
substantial effort to characterize the implicit bias of the optimization
algorithms used, such as gradient descent (GD), and the structural properties
of their preferred solutions. This paper answers an open question in this
literature: For the classification setting, what solution does mirror descent
(MD) converge to? Specifically, motivated by its efficient implementation, we
consider the family of mirror descent algorithms with potential function chosen
as the $p$-th power of the $\ell_p$-norm, which is an important generalization
of GD. We call this algorithm $p$-$\textsf{GD}$. For this family, we
characterize the solutions it obtains and show that it converges in direction
to a generalized maximum-margin solution with respect to the $\ell_p$-norm for
linearly separable classification. While the MD update rule is in general
expensive to compute and perhaps not suitable for deep learning,
$p$-$\textsf{GD}$ is fully parallelizable in the same manner as SGD and can be
used to train deep neural networks with virtually no additional computational
overhead. Using comprehensive experiments with both linear and deep neural
network models, we demonstrate that $p$-$\textsf{GD}$ can noticeably affect the
structure and the generalization performance of the learned models. |
3,952 | Gradient-based explanations for Gaussian Process regression and classification models | Gaussian Processes (GPs) have proven themselves as a reliable and effective
method in probabilistic Machine Learning. Thanks to recent and current
advances, modeling complex data with GPs is becoming more and more feasible.
Thus, these types of models are, nowadays, an interesting alternative to Neural
and Deep Learning methods, which are arguably the current state-of-the-art in
Machine Learning. For the latter, we see an increasing interest in so-called
explainable approaches - in essence methods that aim to make a Machine Learning
model's decision process transparent to humans. Such methods are particularly
needed when illogical or biased reasoning can lead to actual disadvantageous
consequences for humans. Ideally, explainable Machine Learning should help
detect such flaws in a model and aid a subsequent debugging process. One active
line of research in Machine Learning explainability are gradient-based methods,
which have been successfully applied to complex neural networks. Given that GPs
are closed under differentiation, gradient-based explainability for GPs appears
as a promising field of research. This paper is primarily focused on explaining
GP classifiers via gradients where, contrary to GP regression, derivative GPs
are not straightforward to obtain. |
3,953 | Towards Green AI with tensor networks -- Sustainability and innovation enabled by efficient algorithms | The current standard to compare the performance of AI algorithms is mainly
based on one criterion: the model's accuracy. In this context, algorithms with
a higher accuracy (or similar measures) are considered as better. To achieve
new state-of-the-art results, algorithmic development is accompanied by an
exponentially increasing amount of compute. While this has enabled AI research
to achieve remarkable results, AI progress comes at a cost: it is
unsustainable. In this paper, we present a promising tool for sustainable and
thus Green AI: tensor networks (TNs). Being an established tool from
multilinear algebra, TNs have the capability to improve efficiency without
compromising accuracy. Since they can reduce compute significantly, we would
like to highlight their potential for Green AI. We elaborate in both a kernel
machine and deep learning setting how efficiency gains can be achieved with
TNs. Furthermore, we argue that better algorithms should be evaluated in terms
of both accuracy and efficiency. To that end, we discuss different efficiency
criteria and analyze efficiency in an exemplifying experimental setting for
kernel ridge regression. With this paper, we want to raise awareness about
Green AI and showcase its positive impact on sustainability and AI research.
Our key contribution is to demonstrate that TNs enable efficient algorithms and
therefore contribute towards Green AI. In this sense, TNs pave the way for
better algorithms in AI. |
3,954 | Impartial Games: A Challenge for Reinforcement Learning | The AlphaZero algorithm and its successor MuZero have revolutionised several
competitive strategy games, including chess, Go, and shogi and video games like
Atari, by learning to play these games better than any human and any
specialised computer program. Aside from knowing the rules, AlphaZero had no
prior knowledge of each game. This dramatically advanced progress on a
long-standing AI challenge to create programs that can learn for themselves
from first principles.
Theoretically, there are well-known limits to the power of deep learning for
strategy games like chess, Go, and shogi, as they are known to be NEXPTIME
hard. Some papers have argued that the AlphaZero methodology has limitations
and is unsuitable for general AI. However, none of these works has suggested
any specific limits for any particular game.
In this paper, we provide more powerful bottlenecks than previously
suggested. We present the first concrete example of a game - namely the
(children) game of nim - and other impartial games that seem to be a stumbling
block for AlphaZero and similar reinforcement learning algorithms. We show
experimentally that the bottlenecks apply to both the policy and value
networks. Since solving nim can be done in linear time using logarithmic space
i.e. has very low-complexity, our experimental results supersede known
theoretical limits based on many games' PSPACE (and NEXPTIME) completeness.
We show that nim can be learned on small boards, but when the board size
increases, AlphaZero style algorithms rapidly fail to improve.
We quantify the difficulties for various setups, parameter settings and
computational resources. Our results might help expand the AlphaZero self-play
paradigm by allowing it to use meta-actions during training and/or actual game
play like applying abstract transformations, or reading and writing to an
external memory. |
3,955 | TrustGNN: Graph Neural Network based Trust Evaluation via Learnable Propagative and Composable Nature | Trust evaluation is critical for many applications such as cyber security,
social communication and recommender systems. Users and trust relationships
among them can be seen as a graph. Graph neural networks (GNNs) show their
powerful ability for analyzing graph-structural data. Very recently, existing
work attempted to introduce the attributes and asymmetry of edges into GNNs for
trust evaluation, while failed to capture some essential properties (e.g., the
propagative and composable nature) of trust graphs. In this work, we propose a
new GNN based trust evaluation method named TrustGNN, which integrates smartly
the propagative and composable nature of trust graphs into a GNN framework for
better trust evaluation. Specifically, TrustGNN designs specific propagative
patterns for different propagative processes of trust, and distinguishes the
contribution of different propagative processes to create new trust. Thus,
TrustGNN can learn comprehensive node embeddings and predict trust
relationships based on these embeddings. Experiments on some widely-used
real-world datasets indicate that TrustGNN significantly outperforms the
state-of-the-art methods. We further perform analytical experiments to
demonstrate the effectiveness of the key designs in TrustGNN. |
3,956 | Ultra-compact Binary Neural Networks for Human Activity Recognition on RISC-V Processors | Human Activity Recognition (HAR) is a relevant inference task in many mobile
applications. State-of-the-art HAR at the edge is typically achieved with
lightweight machine learning models such as decision trees and Random Forests
(RFs), whereas deep learning is less common due to its high computational
complexity. In this work, we propose a novel implementation of HAR based on
deep neural networks, and precisely on Binary Neural Networks (BNNs), targeting
low-power general purpose processors with a RISC-V instruction set. BNNs yield
very small memory footprints and low inference complexity, thanks to the
replacement of arithmetic operations with bit-wise ones. However, existing BNN
implementations on general purpose processors impose constraints tailored to
complex computer vision tasks, which result in over-parametrized models for
simpler problems like HAR. Therefore, we also introduce a new BNN inference
library, which targets ultra-compact models explicitly. With experiments on a
single-core RISC-V processor, we show that BNNs trained on two HAR datasets
obtain higher classification accuracy compared to a state-of-the-art baseline
based on RFs. Furthermore, our BNN reaches the same accuracy of a RF with
either less memory (up to 91%) or more energy-efficiency (up to 70%), depending
on the complexity of the features extracted by the RF. |
3,957 | Residual-Concatenate Neural Network with Deep Regularization Layers for Binary Classification | Many complex Deep Learning models are used with different variations for
various prognostication tasks. The higher learning parameters not necessarily
ensure great accuracy. This can be solved by considering changes in very deep
models with many regularization based techniques. In this paper we train a deep
neural network that uses many regularization layers with residual and
concatenation process for best fit with Polycystic Ovary Syndrome Diagnosis
prognostication. The network was built with improvements from every step of
failure to meet the needs of the data and achieves an accuracy of 99.3%
seamlessly. |
3,958 | An Experimental Comparison Between Temporal Difference and Residual Gradient with Neural Network Approximation | Gradient descent or its variants are popular in training neural networks.
However, in deep Q-learning with neural network approximation, a type of
reinforcement learning, gradient descent (also known as Residual Gradient (RG))
is barely used to solve Bellman residual minimization problem. On the contrary,
Temporal Difference (TD), an incomplete gradient descent method prevails. In
this work, we perform extensive experiments to show that TD outperforms RG,
that is, when the training leads to a small Bellman residual error, the
solution found by TD has a better policy and is more robust against the
perturbation of neural network parameters. We further use experiments to reveal
a key difference between reinforcement learning and supervised learning, that
is, a small Bellman residual error can correspond to a bad policy in
reinforcement learning while the test loss function in supervised learning is a
standard index to indicate the performance. We also empirically examine that
the missing term in TD is a key reason why RG performs badly. Our work shows
that the performance of a deep Q-learning solution is closely related to the
training dynamics and how an incomplete gradient descent method can find a good
policy is interesting for future study. |
3,959 | Towards Symbolic Time Series Representation Improved by Kernel Density Estimators | This paper deals with symbolic time series representation. It builds up on
the popular mapping technique Symbolic Aggregate approXimation algorithm (SAX),
which is extensively utilized in sequence classification, pattern mining,
anomaly detection, time series indexing and other data mining tasks. However,
the disadvantage of this method is, that it works reliably only for time series
with Gaussian-like distribution. In our previous work we have proposed an
improvement of SAX, called dwSAX, which can deal with Gaussian as well as
non-Gaussian data distribution. Recently we have made further progress in our
solution - edwSAX. Our goal was to optimally cover the information space by
means of sufficient alphabet utilization; and to satisfy lower bounding
criterion as tight as possible. We describe here our approach, including
evaluation on commonly employed tasks such as time series reconstruction error
and Euclidean distance lower bounding with promising improvements over SAX. |
3,960 | An Evolutionary Approach to Dynamic Introduction of Tasks in Large-scale Multitask Learning Systems | Multitask learning assumes that models capable of learning from multiple
tasks can achieve better quality and efficiency via knowledge transfer, a key
feature of human learning. Though, state of the art ML models rely on high
customization for each task and leverage size and data scale rather than
scaling the number of tasks. Also, continual learning, that adds the temporal
aspect to multitask, is often focused to the study of common pitfalls such as
catastrophic forgetting instead of being studied at a large scale as a critical
component to build the next generation artificial intelligence. We propose an
evolutionary method that can generate a large scale multitask model, and can
support the dynamic and continuous addition of new tasks. The generated
multitask model is sparsely activated and integrates a task-based routing that
guarantees bounded compute cost and fewer added parameters per task as the
model expands. The proposed method relies on a knowledge compartmentalization
technique to achieve immunity against catastrophic forgetting and other common
pitfalls such as gradient interference and negative transfer. We empirically
show that the proposed method can jointly solve and achieve competitive results
on 69image classification tasks, for example achieving the best test accuracy
reported fora model trained only on public data for competitive tasks such as
cifar10: 99.43%. |
3,961 | An Empirical Study on Distribution Shift Robustness From the Perspective of Pre-Training and Data Augmentation | The performance of machine learning models under distribution shift has been
the focus of the community in recent years. Most of current methods have been
proposed to improve the robustness to distribution shift from the algorithmic
perspective, i.e., designing better training algorithms to help the
generalization in shifted test distributions. This paper studies the
distribution shift problem from the perspective of pre-training and data
augmentation, two important factors in the practice of deep learning that have
not been systematically investigated by existing work. By evaluating seven
pre-trained models, including ResNets and ViT's with self-supervision and
supervision mode, on five important distribution-shift datasets, from WILDS and
DomainBed benchmarks, with five different learning algorithms, we provide the
first comprehensive empirical study focusing on pre-training and data
augmentation. With our empirical result obtained from 1,330 models, we provide
the following main observations: 1) ERM combined with data augmentation can
achieve state-of-the-art performance if we choose a proper pre-trained model
respecting the data property; 2) specialized algorithms further improve the
robustness on top of ERM when handling a specific type of distribution shift,
e.g., GroupDRO for spurious correlation and CORAL for large-scale
out-of-distribution data; 3) Comparing different pre-training modes,
architectures and data sizes, we provide novel observations about pre-training
on distribution shift, which sheds light on designing or selecting pre-training
strategy for different kinds of distribution shifts. In summary, our empirical
study provides a comprehensive baseline for a wide range of pre-training models
fine-tuned with data augmentation, which potentially inspires research
exploiting the power of pre-training and data augmentation in the future of
distribution shift study. |
3,962 | NECA: Network-Embedded Deep Representation Learning for Categorical Data | We propose NECA, a deep representation learning method for categorical data.
Built upon the foundations of network embedding and deep unsupervised
representation learning, NECA deeply embeds the intrinsic relationship among
attribute values and explicitly expresses data objects with numeric vector
representations. Designed specifically for categorical data, NECA can support
important downstream data mining tasks, such as clustering. Extensive
experimental analysis demonstrated the effectiveness of NECA. |
3,963 | Fast Stochastic Composite Minimization and an Accelerated Frank-Wolfe Algorithm under Parallelization | We consider the problem of minimizing the sum of two convex functions. One of
those functions has Lipschitz-continuous gradients, and can be accessed via
stochastic oracles, whereas the other is "simple". We provide a Bregman-type
algorithm with accelerated convergence in function values to a ball containing
the minimum. The radius of this ball depends on problem-dependent constants,
including the variance of the stochastic oracle. We further show that this
algorithmic setup naturally leads to a variant of Frank-Wolfe achieving
acceleration under parallelization. More precisely, when minimizing a smooth
convex function on a bounded domain, we show that one can achieve an $\epsilon$
primal-dual gap (in expectation) in $\tilde{O}(1/ \sqrt{\epsilon})$ iterations,
by only accessing gradients of the original function and a linear maximization
oracle with $O(1/\sqrt{\epsilon})$ computing units in parallel. We illustrate
this fast convergence on synthetic numerical experiments. |
3,964 | Machine learning method for return direction forecasting of Exchange Traded Funds using classification and regression models | This article aims to propose and apply a machine learning method to analyze
the direction of returns from Exchange Traded Funds (ETFs) using the historical
return data of its components, helping to make investment strategy decisions
through a trading algorithm. In methodological terms, regression and
classification models were applied, using standard datasets from Brazilian and
American markets, in addition to algorithmic error metrics. In terms of
research results, they were analyzed and compared to those of the Na\"ive
forecast and the returns obtained by the buy & hold technique in the same
period of time. In terms of risk and return, the models mostly performed better
than the control metrics, with emphasis on the linear regression model and the
classification models by logistic regression, support vector machine (using the
LinearSVC model), Gaussian Naive Bayes and K-Nearest Neighbors, where in
certain datasets the returns exceeded by two times and the Sharpe ratio by up
to four times those of the buy & hold control model. |
3,965 | Deep interpretable ensembles | Ensembles improve prediction performance and allow uncertainty quantification
by aggregating predictions from multiple models. In deep ensembling, the
individual models are usually black box neural networks, or recently, partially
interpretable semi-structured deep transformation models. However,
interpretability of the ensemble members is generally lost upon aggregation.
This is a crucial drawback of deep ensembles in high-stake decision fields, in
which interpretable models are desired. We propose a novel transformation
ensemble which aggregates probabilistic predictions with the guarantee to
preserve interpretability and yield uniformly better predictions than the
ensemble members on average. Transformation ensembles are tailored towards
interpretable deep transformation models but are applicable to a wider range of
probabilistic neural networks. In experiments on several publicly available
data sets, we demonstrate that transformation ensembles perform on par with
classical deep ensembles in terms of prediction performance, discrimination,
and calibration. In addition, we demonstrate how transformation ensembles
quantify both aleatoric and epistemic uncertainty, and produce minimax optimal
predictions under certain conditions. |
3,966 | Service Discovery in Social Internet of Things using Graph Neural Networks | Internet-of-Things (IoT) networks intelligently connect thousands of physical
entities to provide various services for the community. It is witnessing an
exponential expansion, which is complicating the process of discovering IoT
devices existing in the network and requesting corresponding services from
them. As the highly dynamic nature of the IoT environment hinders the use of
traditional solutions of service discovery, we aim, in this paper, to address
this issue by proposing a scalable resource allocation neural model adequate
for heterogeneous large-scale IoT networks. We devise a Graph Neural Network
(GNN) approach that utilizes the social relationships formed between the
devices in the IoT network to reduce the search space of any entity lookup and
acquire a service from another device in the network. This proposed resource
allocation approach surpasses standardization issues and embeds the structure
and characteristics of the social IoT graph, by the means of GNNs, for eventual
clustering analysis process. Simulation results applied on a real-world dataset
illustrate the performance of this solution and its significant efficiency to
operate on large-scale IoT networks. |
3,967 | VeriFi: Towards Verifiable Federated Unlearning | Federated learning (FL) is a collaborative learning paradigm where
participants jointly train a powerful model without sharing their private data.
One desirable property for FL is the implementation of the right to be
forgotten (RTBF), i.e., a leaving participant has the right to request to
delete its private data from the global model. However, unlearning itself may
not be enough to implement RTBF unless the unlearning effect can be
independently verified, an important aspect that has been overlooked in the
current literature. In this paper, we prompt the concept of verifiable
federated unlearning, and propose VeriFi, a unified framework integrating
federated unlearning and verification that allows systematic analysis of the
unlearning and quantification of its effect, with different combinations of
multiple unlearning and verification methods. In VeriFi, the leaving
participant is granted the right to verify (RTV), that is, the participant
notifies the server before leaving, then actively verifies the unlearning
effect in the next few communication rounds. The unlearning is done at the
server side immediately after receiving the leaving notification, while the
verification is done locally by the leaving participant via two steps: marking
(injecting carefully-designed markers to fingerprint the leaver) and checking
(examining the change of the global model's performance on the markers). Based
on VeriFi, we conduct the first systematic and large-scale study for verifiable
federated unlearning, considering 7 unlearning methods and 5 verification
methods. Particularly, we propose a more efficient and FL-friendly unlearning
method, and two more effective and robust non-invasive-verification methods. We
extensively evaluate VeriFi on 7 datasets and 4 types of deep learning models.
Our analysis establishes important empirical understandings for more
trustworthy federated unlearning. |
3,968 | Scalable Online Change Detection for High-dimensional Data Streams | Detecting changes in data streams is a core objective in their analysis and
has applications in, say, predictive maintenance, fraud detection, and
medicine. A principled approach to detect changes is to compare distributions
observed within the stream to each other. However, data streams often are
high-dimensional, and changes can be complex, e.g., only manifest themselves in
higher moments. The streaming setting also imposes heavy memory and computation
restrictions. We propose an algorithm, Maximum Mean Discrepancy Adaptive
Windowing (MMDAW), which leverages the well-known Maximum Mean Discrepancy
(MMD) two-sample test, and facilitates its efficient online computation on
windows whose size it flexibly adapts. As MMD is sensitive to any change in the
underlying distribution, our algorithm is a general-purpose non-parametric
change detector that fulfills the requirements imposed by the streaming
setting. Our experiments show that MMDAW achieves better detection quality than
state-of-the-art competitors. |
3,969 | Eliciting Transferability in Multi-task Learning with Task-level Mixture-of-Experts | Recent work suggests that transformer models are capable of multi-task
learning on diverse NLP tasks. However, the potential of these models may be
limited as they use the same set of parameters for all tasks. In contrast,
humans tackle tasks in a more flexible way, by making proper presumptions on
what skills and knowledge are relevant and executing only the necessary
computations. Inspired by this, we propose to use task-level mixture-of-expert
models, which has a collection of transformer layers (i.e., experts) and a
router component to choose among these experts dynamically and flexibly. We
show that the learned routing decisions and experts partially rediscover human
categorization of NLP tasks -- certain experts are strongly associated with
extractive tasks, some with classification tasks, and some with tasks requiring
world knowledge. |
3,970 | Surprises in adversarially-trained linear regression | State-of-the-art machine learning models can be vulnerable to very small
input perturbations that are adversarially constructed. Adversarial training is
one of the most effective approaches to defend against such examples. We show
that for linear regression problems, adversarial training can be formulated as
a convex problem. This fact is then used to show that $\ell_\infty$-adversarial
training produces sparse solutions and has many similarities to the lasso
method. Similarly, $\ell_2$-adversarial training has similarities with ridge
regression. We use a robust regression framework to analyze and understand
these similarities and also point to some differences. Finally, we show how
adversarial training behaves differently from other regularization methods when
estimating overparameterized models (i.e., models with more parameters than
datapoints). It minimizes a sum of three terms which regularizes the solution,
but unlike lasso and ridge regression, it can sharply transition into an
interpolation mode. We show that for sufficiently many features or sufficiently
small regularization parameters, the learned model perfectly interpolates the
training data while still exhibiting good out-of-sample performance. |
3,971 | Train Flat, Then Compress: Sharpness-Aware Minimization Learns More Compressible Models | Model compression by way of parameter pruning, quantization, or distillation
has recently gained popularity as an approach for reducing the computational
requirements of modern deep neural network models for NLP. Pruning unnecessary
parameters has emerged as a simple and effective method for compressing large
models that is compatible with a wide variety of contemporary off-the-shelf
hardware (unlike quantization), and that requires little additional training
(unlike distillation). Pruning approaches typically take a large, accurate
model as input, then attempt to discover a smaller subnetwork of that model
capable of achieving end-task accuracy comparable to the full model. Inspired
by previous work suggesting a connection between simpler, more generalizable
models and those that lie within flat basins in the loss landscape, we propose
to directly optimize for flat minima while performing task-specific pruning,
which we hypothesize should lead to simpler parameterizations and thus more
compressible models. In experiments combining sharpness-aware minimization with
both iterative magnitude pruning and structured pruning approaches, we show
that optimizing for flat minima consistently leads to greater compressibility
of parameters compared to standard Adam optimization when fine-tuning BERT
models, leading to higher rates of compression with little to no loss in
accuracy on the GLUE classification benchmark. |
3,972 | Ground-Truth Labels Matter: A Deeper Look into Input-Label Demonstrations | Despite recent explosion in research interests, in-context learning and the
precise impact of the quality of demonstrations remain elusive. While, based on
current literature, it is expected that in-context learning shares a similar
mechanism to supervised learning, Min et al. (2022) recently reported that,
surprisingly, input-label correspondence is less important than other aspects
of prompt demonstrations. Inspired by this counter-intuitive observation, we
re-examine the importance of ground truth labels on in-context learning from
diverse and statistical points of view. With the aid of the newly introduced
metrics, i.e., Ground-truth Label Effect Ratio (GLER), demo-gain, and label
sensitivity, we find that the impact of the correct input-label matching can
vary according to different configurations. Expanding upon the previous key
finding on the role of demonstrations, the complementary and contrastive
results suggest that one might need to take more care when estimating the
impact of each component in in-context learning demonstrations. |
3,973 | Rethinking Fano's Inequality in Ensemble Learning | We propose a fundamental theory on ensemble learning that evaluates a given
ensemble system by a well-grounded set of metrics. Previous studies used a
variant of Fano's inequality of information theory and derived a lower bound of
the classification error rate on the basis of the accuracy and diversity of
models. We revisit the original Fano's inequality and argue that the studies
did not take into account the information lost when multiple model predictions
are combined into a final prediction. To address this issue, we generalize the
previous theory to incorporate the information loss. Further, we empirically
validate and demonstrate the proposed theory through extensive experiments on
actual systems. The theory reveals the strengths and weaknesses of systems on
each metric, which will push the theoretical understanding of ensemble learning
and give us insights into designing systems. |
3,974 | Training Language Models with Memory Augmentation | Recent work has improved language models remarkably by equipping them with a
non-parametric memory component. However, most existing approaches only
introduce memories at testing time, or represent them using a separately
trained encoder -- resulting in sub-optimal training of the language model. In
this work, we present TRIME, a novel yet simple training approach designed for
training language models with memory augmentation. Our approach uses a training
objective that directly takes in-batch examples as accessible memory. We also
present new methods for memory construction and data batching, which are used
for adapting to different sets of memories -- local, long-term, and external
memory -- at testing time. We evaluate our approach on multiple language
modeling and machine translation benchmarks. We find that simply replacing the
vanilla language modeling objective by ours greatly reduces the perplexity,
without modifying the model architecture or incorporating extra context (e.g.,
18.70 $\to$ 17.76 on WikiText-103). We further augment language models with
long-range contexts and external knowledge and demonstrate significant gains
over previous memory-augmented approaches. |
3,975 | MAPLE-X: Latency Prediction with Explicit Microprocessor Prior Knowledge | Deep neural network (DNN) latency characterization is a time-consuming
process and adds significant cost to Neural Architecture Search (NAS) processes
when searching for efficient convolutional neural networks for embedded vision
applications. DNN Latency is a hardware dependent metric and requires direct
measurement or inference on target hardware. A recently introduced latency
estimation technique known as MAPLE predicts DNN execution time on previously
unseen hardware devices by using hardware performance counters. Leveraging
these hardware counters in the form of an implicit prior, MAPLE achieves
state-of-the-art performance in latency prediction. Here, we propose MAPLE-X
which extends MAPLE by incorporating explicit prior knowledge of hardware
devices and DNN architecture latency to better account for model stability and
robustness. First, by identifying DNN architectures that exhibit a similar
latency to each other, we can generate multiple virtual examples to
significantly improve the accuracy over MAPLE. Secondly, the hardware
specifications are used to determine the similarity between training and test
hardware to emphasize training samples captured from comparable devices
(domains) and encourages improved domain alignment. Experimental results using
a convolution neural network NAS benchmark across different types of devices,
including an Intel processor that is now used for embedded vision applications,
demonstrate a 5% improvement over MAPLE and 9% over HELP. Furthermore, we
include ablation studies to independently assess the benefits of virtual
examples and hardware-based sample importance. |
3,976 | Fast Inference and Transfer of Compositional Task Structures for Few-shot Task Generalization | We tackle real-world problems with complex structures beyond the pixel-based
game or simulator. We formulate it as a few-shot reinforcement learning problem
where a task is characterized by a subtask graph that defines a set of subtasks
and their dependencies that are unknown to the agent. Different from the
previous meta-rl methods trying to directly infer the unstructured task
embedding, our multi-task subtask graph inferencer (MTSGI) first infers the
common high-level task structure in terms of the subtask graph from the
training tasks, and use it as a prior to improve the task inference in testing.
Our experiment results on 2D grid-world and complex web navigation domains show
that the proposed method can learn and leverage the common underlying structure
of the tasks for faster adaptation to the unseen tasks than various existing
algorithms such as meta reinforcement learning, hierarchical reinforcement
learning, and other heuristic agents. |
3,977 | On the Interpretability of Regularisation for Neural Networks Through Model Gradient Similarity | Most complex machine learning and modelling techniques are prone to
over-fitting and may subsequently generalise poorly to future data. Artificial
neural networks are no different in this regard and, despite having a level of
implicit regularisation when trained with gradient descent, often require the
aid of explicit regularisers. We introduce a new framework, Model Gradient
Similarity (MGS), that (1) serves as a metric of regularisation, which can be
used to monitor neural network training, (2) adds insight into how explicit
regularisers, while derived from widely different principles, operate via the
same mechanism underneath by increasing MGS, and (3) provides the basis for a
new regularisation scheme which exhibits excellent performance, especially in
challenging settings such as high levels of label noise or limited sample
sizes. |
3,978 | Lifelong Learning Natural Language Processing Approach for Multilingual Data Classification | The abundance of information in digital media, which in today's world is the
main source of knowledge about current events for the masses, makes it possible
to spread disinformation on a larger scale than ever before. Consequently,
there is a need to develop novel fake news detection approaches capable of
adapting to changing factual contexts and generalizing previously or
concurrently acquired knowledge. To deal with this problem, we propose a
lifelong learning-inspired approach, which allows for fake news detection in
multiple languages and the mutual transfer of knowledge acquired in each of
them. Both classical feature extractors, such as Term frequency-inverse
document frequency or Latent Dirichlet Allocation, and integrated deep NLP
(Natural Language Processing) BERT (Bidirectional Encoder Representations from
Transformers) models paired with MLP (Multilayer Perceptron) classifier, were
employed. The results of experiments conducted on two datasets dedicated to the
fake news classification task (in English and Spanish, respectively), supported
by statistical analysis, confirmed that utilization of additional languages
could improve performance for traditional methods. Also, in some cases
supplementing the deep learning method with classical ones can positively
impact obtained results. The ability of models to generalize the knowledge
acquired between the analyzed languages was also observed. |
3,979 | Autoformalization with Large Language Models | Autoformalization is the process of automatically translating from natural
language mathematics to formal specifications and proofs. A successful
autoformalization system could advance the fields of formal verification,
program synthesis, and artificial intelligence. While the long-term goal of
autoformalization seemed elusive for a long time, we show large language models
provide new prospects towards this goal. We make the surprising observation
that LLMs can correctly translate a significant portion ($25.3\%$) of
mathematical competition problems perfectly to formal specifications in
Isabelle/HOL. We demonstrate the usefulness of this process by improving a
previously introduced neural theorem prover via training on these
autoformalized theorems. Our methodology results in a new state-of-the-art
result on the MiniF2F theorem proving benchmark, improving the proof rate from
$29.6\%$ to $35.2\%$. |
3,980 | Deep Aesthetic Assessment and Retrieval of Breast Cancer Treatment Outcomes | Treatments for breast cancer have continued to evolve and improve in recent
years, resulting in a substantial increase in survival rates, with
approximately 80\% of patients having a 10-year survival period. Given the
serious impact that breast cancer treatments can have on a patient's body
image, consequently affecting her self-confidence and sexual and intimate
relationships, it is paramount to ensure that women receive the treatment that
optimizes both survival and aesthetic outcomes. Currently, there is no gold
standard for evaluating the aesthetic outcome of breast cancer treatment. In
addition, there is no standard way to show patients the potential outcome of
surgery. The presentation of similar cases from the past would be extremely
important to manage women's expectations of the possible outcome. In this work,
we propose a deep neural network to perform the aesthetic evaluation. As a
proof-of-concept, we focus on a binary aesthetic evaluation. Besides its use
for classification, this deep neural network can also be used to find the most
similar past cases by searching for nearest neighbours in the highly semantic
space before classification. We performed the experiments on a dataset
consisting of 143 photos of women after conservative treatment for breast
cancer. The results for accuracy and balanced accuracy showed the superior
performance of our proposed model compared to the state of the art in aesthetic
evaluation of breast cancer treatments. In addition, the model showed a good
ability to retrieve similar previous cases, with the retrieved cases having the
same or adjacent class (in the 4-class setting) and having similar types of
asymmetry. Finally, a qualitative interpretability assessment was also
performed to analyse the robustness and trustworthiness of the model. |
3,981 | Mutual Information Divergence: A Unified Metric for Multimodal Generative Models | Text-to-image generation and image captioning are recently emerged as a new
experimental paradigm to assess machine intelligence. They predict continuous
quantity accompanied by their sampling techniques in the generation, making
evaluation complicated and intractable to get marginal distributions. Based on
a recent trend that multimodal generative evaluations exploit a
vison-and-language pre-trained model, we propose the negative Gaussian
cross-mutual information using the CLIP features as a unified metric, coined by
Mutual Information Divergence (MID). To validate, we extensively compare it
with competing metrics using carefully-generated or human-annotated judgments
in text-to-image generation and image captioning tasks. The proposed MID
significantly outperforms the competitive methods by having consistency across
benchmarks, sample parsimony, and robustness toward the exploited CLIP model.
We look forward to seeing the underrepresented implications of the Gaussian
cross-mutual information in multimodal representation learning and the future
works based on this novel proposition. |
3,982 | Learning Distributions by Generative Adversarial Networks: Approximation and Generalization | We study how well generative adversarial networks (GAN) learn probability
distributions from finite samples by analyzing the convergence rates of these
models. Our analysis is based on a new oracle inequality that decomposes the
estimation error of GAN into the discriminator and generator approximation
errors, generalization error and optimization error. To estimate the
discriminator approximation error, we establish error bounds on approximating
H\"older functions by ReLU neural networks, with explicit upper bounds on the
Lipschitz constant of the network or norm constraint on the weights. For
generator approximation error, we show that neural network can approximately
transform a low-dimensional source distribution to a high-dimensional target
distribution and bound such approximation error by the width and depth of
neural network. Combining the approximation results with generalization bounds
of neural networks from statistical learning theory, we establish the
convergence rates of GANs in various settings, when the error is measured by a
collection of integral probability metrics defined through H\"older classes,
including the Wasserstein distance as a special case. In particular, for
distributions concentrated around a low-dimensional set, we show that the
convergence rates of GANs do not depend on the high ambient dimension, but on
the lower intrinsic dimension. |
3,983 | ORCA: Interpreting Prompted Language Models via Locating Supporting Data Evidence in the Ocean of Pretraining Data | Large pretrained language models have been performing increasingly well in a
variety of downstream tasks via prompting. However, it remains unclear from
where the model learns the task-specific knowledge, especially in a zero-shot
setup. In this work, we want to find evidence of the model's task-specific
competence from pretraining and are specifically interested in locating a very
small subset of pretraining data that directly supports the model in the task.
We call such a subset supporting data evidence and propose a novel method ORCA
to effectively identify it, by iteratively using gradient information related
to the downstream task. This supporting data evidence offers interesting
insights about the prompted language models: in the tasks of sentiment analysis
and textual entailment, BERT shows a substantial reliance on BookCorpus, the
smaller corpus of BERT's two pretraining corpora, as well as on pretraining
examples that mask out synonyms to the task verbalizers. |
3,984 | RobustLR: Evaluating Robustness to Logical Perturbation in Deductive Reasoning | Transformers have been shown to be able to perform deductive reasoning on a
logical rulebase containing rules and statements written in English natural
language. While the progress is promising, it is currently unclear if these
models indeed perform logical reasoning by understanding the underlying logical
semantics in the language. To this end, we propose RobustLR, a suite of
evaluation datasets that evaluate the robustness of these models to minimal
logical edits in rulebases and some standard logical equivalence conditions. In
our experiments with RoBERTa and T5, we find that the models trained in prior
works do not perform consistently on the different perturbations in RobustLR,
thus showing that the models are not robust to the proposed logical
perturbations. Further, we find that the models find it especially hard to
learn logical negation and disjunction operators. Overall, using our evaluation
sets, we demonstrate some shortcomings of the deductive reasoning-based
language models, which can eventually help towards designing better models for
logical reasoning over natural language. |
3,985 | Heterogeneous Reservoir Computing Models for Persian Speech Recognition | Over the last decade, deep-learning methods have been gradually incorporated
into conventional automatic speech recognition (ASR) frameworks to create
acoustic, pronunciation, and language models. Although it led to significant
improvements in ASRs' recognition accuracy, due to their hard constraints
related to hardware requirements (e.g., computing power and memory usage), it
is unclear if such approaches are the most computationally- and
energy-efficient options for embedded ASR applications. Reservoir computing
(RC) models (e.g., echo state networks (ESNs) and liquid state machines
(LSMs)), on the other hand, have been proven inexpensive to train, have vastly
fewer parameters, and are compatible with emergent hardware technologies.
However, their performance in speech processing tasks is relatively inferior to
that of the deep-learning-based models. To enhance the accuracy of the RC in
ASR applications, we propose heterogeneous single and multi-layer ESNs to
create non-linear transformations of the inputs that capture temporal context
at different scales. To test our models, we performed a speech recognition task
on the Farsdat Persian dataset. Since, to the best of our knowledge, standard
RC has not yet been employed to conduct any Persian ASR tasks, we also trained
conventional single-layer and deep ESNs to provide baselines for comparison.
Besides, we compared the RC performance with a standard long-short-term memory
(LSTM) model. Heterogeneous RC models (1) show improved performance to the
standard RC models; (2) perform on par in terms of recognition accuracy with
the LSTM, and (3) reduce the training time considerably. |
3,986 | Scalable Multi-Agent Model-Based Reinforcement Learning | Recent Multi-Agent Reinforcement Learning (MARL) literature has been largely
focused on Centralized Training with Decentralized Execution (CTDE) paradigm.
CTDE has been a dominant approach for both cooperative and mixed environments
due to its capability to efficiently train decentralized policies. While in
mixed environments full autonomy of the agents can be a desirable outcome,
cooperative environments allow agents to share information to facilitate
coordination. Approaches that leverage this technique are usually referred as
communication methods, as full autonomy of agents is compromised for better
performance. Although communication approaches have shown impressive results,
they do not fully leverage this additional information during training phase.
In this paper, we propose a new method called MAMBA which utilizes Model-Based
Reinforcement Learning (MBRL) to further leverage centralized training in
cooperative environments. We argue that communication between agents is enough
to sustain a world model for each agent during execution phase while imaginary
rollouts can be used for training, removing the necessity to interact with the
environment. These properties yield sample efficient algorithm that can scale
gracefully with the number of agents. We empirically confirm that MAMBA
achieves good performance while reducing the number of interactions with the
environment up to an orders of magnitude compared to Model-Free
state-of-the-art approaches in challenging domains of SMAC and Flatland. |
3,987 | Towards a Fair Comparison and Realistic Design and Evaluation Framework of Android Malware Detectors | As in other cybersecurity areas, machine learning (ML) techniques have
emerged as a promising solution to detect Android malware. In this sense, many
proposals employing a variety of algorithms and feature sets have been
presented to date, often reporting impresive detection performances. However,
the lack of reproducibility and the absence of a standard evaluation framework
make these proposals difficult to compare. In this paper, we perform an
analysis of 10 influential research works on Android malware detection using a
common evaluation framework. We have identified five factors that, if not taken
into account when creating datasets and designing detectors, significantly
affect the trained ML models and their performances. In particular, we analyze
the effect of (1) the presence of duplicated samples, (2) label
(goodware/greyware/malware) attribution, (3) class imbalance, (4) the presence
of apps that use evasion techniques and, (5) the evolution of apps. Based on
this extensive experimentation, we conclude that the studied ML-based detectors
have been evaluated optimistically, which justifies the good published results.
Our findings also highlight that it is imperative to generate realistic
datasets, taking into account the factors mentioned above, to enable the design
and evaluation of better solutions for Android malware detection. |
3,988 | Learning dynamics from partial observations with structured neural ODEs | Identifying dynamical systems from experimental data is a notably difficult
task. Prior knowledge generally helps, but the extent of this knowledge varies
with the application, and customized models are often needed. We propose a
flexible framework to incorporate a broad spectrum of physical insight into
neural ODE-based system identification, giving physical interpretability to the
resulting latent space. This insight is either enforced through hard
constraints in the optimization problem or added in its cost function. In order
to link the partial and possibly noisy observations to the latent state, we
rely on tools from nonlinear observer theory to build a recognition model. We
demonstrate the performance of the proposed approach on numerical simulations
and on an experimental dataset from a robotic exoskeleton. |
3,989 | Learning from time-dependent streaming data with online stochastic algorithms | We study stochastic algorithms in a streaming framework, trained on samples
coming from a dependent data source. In this streaming framework, we analyze
the convergence of Stochastic Gradient (SG) methods in a non-asymptotic manner;
this includes various SG methods such as the well-known stochastic gradient
descent (i.e., Robbins-Monro algorithm), mini-batch SG methods, together with
their averaged estimates (i.e., Polyak-Ruppert averaged). Our results form a
heuristic by linking the level of dependency and convexity to the rest of the
model parameters. This heuristic provides new insights into choosing the
optimal learning rate, which can help increase the stability of SGbased
methods; these investigations suggest large streaming batches with slow
decaying learning rates for highly dependent data sources. |
3,990 | RLPrompt: Optimizing Discrete Text Prompts With Reinforcement Learning | Prompting has shown impressive success in enabling large pretrained language
models (LMs) to perform diverse NLP tasks, especially when only few downstream
data are available. Automatically finding the optimal prompt for each task,
however, is challenging. Most existing work resorts to tuning soft prompt
(e.g., embeddings) which falls short of interpretability, reusability across
LMs, and applicability when gradients are not accessible. Discrete prompt, on
the other hand, is difficult to optimize, and is often created by "enumeration
(e.g., paraphrasing)-then-selection" heuristics that do not explore the prompt
space systematically. This paper proposes RLPrompt, an efficient discrete
prompt optimization approach with reinforcement learning (RL). RLPrompt
formulates a parameter-efficient policy network that generates the desired
discrete prompt after training with reward. To overcome the complexity and
stochasticity of reward signals by the large LM environment, we incorporate
effective reward stabilization that substantially enhances the training
efficiency. RLPrompt is flexibly applicable to different types of LMs, such as
masked (e.g., BERT) and left-to-right models (e.g., GPTs), for both
classification and generation tasks. Experiments on few-shot classification and
unsupervised text style transfer show superior performance over a wide range of
existing finetuning or prompting methods. Interestingly, the resulting
optimized prompts are often ungrammatical gibberish text; and surprisingly,
those gibberish prompts are transferrable between different LMs to retain
significant performance, indicating LM prompting may not follow human language
patterns. |
3,991 | Misleading Deep-Fake Detection with GAN Fingerprints | Generative adversarial networks (GANs) have made remarkable progress in
synthesizing realistic-looking images that effectively outsmart even humans.
Although several detection methods can recognize these deep fakes by checking
for image artifacts from the generation process, multiple counterattacks have
demonstrated their limitations. These attacks, however, still require certain
conditions to hold, such as interacting with the detection method or adjusting
the GAN directly. In this paper, we introduce a novel class of simple
counterattacks that overcomes these limitations. In particular, we show that an
adversary can remove indicative artifacts, the GAN fingerprint, directly from
the frequency spectrum of a generated image. We explore different realizations
of this removal, ranging from filtering high frequencies to more nuanced
frequency-peak cleansing. We evaluate the performance of our attack with
different detection methods, GAN architectures, and datasets. Our results show
that an adversary can often remove GAN fingerprints and thus evade the
detection of generated images. |
3,992 | Is a Question Decomposition Unit All We Need? | Large Language Models (LMs) have achieved state-of-the-art performance on
many Natural Language Processing (NLP) benchmarks. With the growing number of
new benchmarks, we build bigger and more complex LMs. However, building new LMs
may not be an ideal option owing to the cost, time and environmental impact
associated with it. We explore an alternative route: can we modify data by
expressing it in terms of the model's strengths, so that a question becomes
easier for models to answer? We investigate if humans can decompose a hard
question into a set of simpler questions that are relatively easier for models
to solve. We analyze a range of datasets involving various forms of reasoning
and find that it is indeed possible to significantly improve model performance
(24% for GPT3 and 29% for RoBERTa-SQuAD along with a symbolic calculator) via
decomposition. Our approach provides a viable option to involve people in NLP
research in a meaningful way. Our findings indicate that Human-in-the-loop
Question Decomposition (HQD) can potentially provide an alternate path to
building large LMs. |
3,993 | A Rotated Hyperbolic Wrapped Normal Distribution for Hierarchical Representation Learning | We present a rotated hyperbolic wrapped normal distribution (RoWN), a simple
yet effective alteration of a hyperbolic wrapped normal distribution (HWN). The
HWN expands the domain of probabilistic modeling from Euclidean to hyperbolic
space, where a tree can be embedded with arbitrary low distortion in theory. In
this work, we analyze the geometric properties of the diagonal HWN, a standard
choice of distribution in probabilistic modeling. The analysis shows that the
distribution is inappropriate to represent the data points at the same
hierarchy level through their angular distance with the same norm in the
Poincar\'e disk model. We then empirically verify the presence of limitations
of HWN, and show how RoWN, the newly proposed distribution, can alleviate the
limitations on various hierarchical datasets, including noisy synthetic binary
tree, WordNet, and Atari 2600 Breakout. |
3,994 | Structured Uncertainty in the Observation Space of Variational Autoencoders | Variational autoencoders (VAEs) are a popular class of deep generative models
with many variants and a wide range of applications. Improvements upon the
standard VAE mostly focus on the modelling of the posterior distribution over
the latent space and the properties of the neural network decoder. In contrast,
improving the model for the observational distribution is rarely considered and
typically defaults to a pixel-wise independent categorical or normal
distribution. In image synthesis, sampling from such distributions produces
spatially-incoherent results with uncorrelated pixel noise, resulting in only
the sample mean being somewhat useful as an output prediction. In this paper,
we aim to stay true to VAE theory by improving the samples from the
observational distribution. We propose an alternative model for the observation
space, encoding spatial dependencies via a low-rank parameterisation. We
demonstrate that this new observational distribution has the ability to capture
relevant covariance between pixels, resulting in spatially-coherent samples. In
contrast to pixel-wise independent distributions, our samples seem to contain
semantically meaningful variations from the mean allowing the prediction of
multiple plausible outputs with a single forward pass. |
3,995 | Skill Machines: Temporal Logic Composition in Reinforcement Learning | A major challenge in reinforcement learning is specifying tasks in a manner
that is both interpretable and verifiable. One common approach is to specify
tasks through reward machines -- finite state machines that encode the task to
be solved. We introduce skill machines, a representation that can be learned
directly from these reward machines that encode the solution to such tasks. We
propose a framework where an agent first learns a set of base skills in a
reward-free setting, and then combines these skills with the learned skill
machine to produce composite behaviours specified by any regular language, such
as linear temporal logics. This provides the agent with the ability to map from
complex logical task specifications to near-optimal behaviours zero-shot. We
demonstrate our approach in both a tabular and high-dimensional video game
environment, where an agent is faced with several of these complex,
long-horizon tasks. Our results indicate that the agent is capable of
satisfying extremely complex task specifications, producing near optimal
performance with no further learning. Finally, we demonstrate that the
performance of skill machines can be improved with regular offline
reinforcement learning algorithms when optimal behaviours are desired. |
3,996 | Toward Discovering Options that Achieve Faster Planning | We propose a new objective for option discovery that emphasizes the
computational advantage of using options in planning. For a given set of
episodic tasks and a given number of options, the objective prefers options
that can be used to achieve a high return by composing few options. By
composing few options, fast planning can be achieved. When faced with new tasks
similar to the given ones, the discovered options are also expected to
accelerate planning. Our objective extends the objective proposed by Harb et
al. (2018) for the single-task setting to the multi-task setting. A closer look
at Harb et al.'s objective shows that the best options discovered given one
task are not likely to be useful for future unseen tasks and that the
multi-task setting is indeed necessary for this purpose. In the same paper,
Harb et al. also proposed an algorithm to optimize their objective, and the
algorithm can be naturally extended to the multi-task setting. We empirically
show that in the four-room domain the extension does not achieve a high
objective value and propose a new algorithm that better optimizes the proposed
objective. In the same four-room domain, we show that 1) a higher objective
value is typically associated with options with which fewer planning iterations
are needed to achieve near-optimal performance, 2) our new algorithm achieves a
high objective value, which is close to the value achieved by a set of
human-designed options, 3) the best number of planning iterations given the
discovered options is much smaller and matches it obtained given human-designed
options, and 4) the options produced by our algorithm also make intuitive sense
because they move to and terminate at cells near hallways connecting two
neighbor rooms. |
3,997 | Exact Phase Transitions in Deep Learning | This work reports deep-learning-unique first-order and second-order phase
transitions, whose phenomenology closely follows that in statistical physics.
In particular, we prove that the competition between prediction error and model
complexity in the training loss leads to the second-order phase transition for
nets with one hidden layer and the first-order phase transition for nets with
more than one hidden layer. The proposed theory is directly relevant to the
optimization of neural networks and points to an origin of the posterior
collapse problem in Bayesian deep learning. |
3,998 | Memorization in NLP Fine-tuning Methods | Large language models are shown to present privacy risks through memorization
of training data, and several recent works have studied such risks for the
pre-training phase. Little attention, however, has been given to the
fine-tuning phase and it is not well understood how different fine-tuning
methods (such as fine-tuning the full model, the model head, and adapter)
compare in terms of memorization risk. This presents increasing concern as the
"pre-train and fine-tune" paradigm proliferates. In this paper, we empirically
study memorization of fine-tuning methods using membership inference and
extraction attacks, and show that their susceptibility to attacks is very
different. We observe that fine-tuning the head of the model has the highest
susceptibility to attacks, whereas fine-tuning smaller adapters appears to be
less vulnerable to known extraction attacks. |
3,999 | The Dialog Must Go On: Improving Visual Dialog via Generative Self-Training | Visual dialog (VisDial) is a task of answering a sequence of questions
grounded in an image, using the dialog history as context. Prior work has
trained the dialog agents solely on VisDial data via supervised learning or
leveraged pre-training on related vision-and-language datasets. This paper
presents a semi-supervised learning approach for visually-grounded dialog,
called Generative Self-Training (GST), to leverage unlabeled images on the Web.
Specifically, GST first retrieves in-domain images through out-of-distribution
detection and generates synthetic dialogs regarding the images via multimodal
conditional text generation. GST then trains a dialog agent on the synthetic
and the original VisDial data. As a result, GST scales the amount of training
data up to an order of magnitude that of VisDial (1.2M to 12.9M QA data). For
robust training of the generated dialogs, we also propose perplexity-based data
selection and multimodal consistency regularization. Evaluation on VisDial v1.0
and v0.9 datasets shows that GST achieves new state-of-the-art results on both
datasets. We further observe strong performance gains in the low-data regime
(up to 9.35 absolute points on NDCG). |
Subsets and Splits
No saved queries yet
Save your SQL queries to embed, download, and access them later. Queries will appear here once saved.