Unnamed: 0
int64 0
5k
| title
stringlengths 9
210
| abstract
stringlengths 164
1.92k
|
|---|---|---|
2,800 | Using Connectome Features to Constrain Echo State Networks | We report an improvement to the conventional Echo State Network (ESN), which
already achieves competitive performance in one-dimensional time series
prediction of dynamical systems. Our model -- a 20$\%$-dense ESN with reservoir
weights derived from a fruit fly connectome (and from its bootstrapped
distribution) -- yields superior performance on a chaotic time series
prediction task, and furthermore alleviates the ESN's high-variance problem. We
also find that an arbitrary positioning of weights can degrade ESN performance
and variance; and that this can be remedied in particular by employing
connectome-derived weight positions. Herein we consider four connectome
features -- namely, the sparsity, positioning, distribution, and clustering of
weights -- and construct corresponding model classes (A, B, B${}_2$, C) from an
appropriate null model ESN; one with its reservoir layer replaced by a fruit
fly connectivity matrix. After tuning relevant hyperparameters and selecting
the best instance of each model class, we train and validate all models for
multi-step prediction on size-variants (50, 250, 500, and 750 training input
steps) of the Mackey-Glass chaotic time series; and compute their performance
(Mean-Squared Error) and variance across train-validate trials. |
2,801 | Bandit Theory and Thompson Sampling-Guided Directed Evolution for Sequence Optimization | Directed Evolution (DE), a landmark wet-lab method originated in 1960s,
enables discovery of novel protein designs via evolving a population of
candidate sequences. Recent advances in biotechnology has made it possible to
collect high-throughput data, allowing the use of machine learning to map out a
protein's sequence-to-function relation. There is a growing interest in machine
learning-assisted DE for accelerating protein optimization. Yet the theoretical
understanding of DE, as well as the use of machine learning in DE, remains
limited. In this paper, we connect DE with the bandit learning theory and make
a first attempt to study regret minimization in DE. We propose a Thompson
Sampling-guided Directed Evolution (TS-DE) framework for sequence optimization,
where the sequence-to-function mapping is unknown and querying a single value
is subject to costly and noisy measurements. TS-DE updates a posterior of the
function based on collected measurements. It uses a posterior-sampled function
estimate to guide the crossover recombination and mutation steps in DE. In the
case of a linear model, we show that TS-DE enjoys a Bayesian regret of order
$\tilde O(d^{2}\sqrt{MT})$, where $d$ is feature dimension, $M$ is population
size and $T$ is number of rounds. This regret bound is nearly optimal,
confirming that bandit learning can provably accelerate DE. It may have
implications for more general sequence optimization and evolutionary
algorithms. |
2,802 | Inference for Interpretable Machine Learning: Fast, Model-Agnostic Confidence Intervals for Feature Importance | In order to trust machine learning for high-stakes problems, we need models
to be both reliable and interpretable. Recently, there has been a growing body
of work on interpretable machine learning which generates human understandable
insights into data, models, or predictions. At the same time, there has been
increased interest in quantifying the reliability and uncertainty of machine
learning predictions, often in the form of confidence intervals for predictions
using conformal inference. Yet, there has been relatively little attention
given to the reliability and uncertainty of machine learning interpretations,
which is the focus of this paper. Our goal is to develop confidence intervals
for a widely-used form of machine learning interpretation: feature importance.
We specifically seek to develop universal model-agnostic and assumption-light
confidence intervals for feature importance that will be valid for any machine
learning model and for any regression or classification task. We do so by
leveraging a form of random observation and feature subsampling called
minipatch ensembles and show that our approach provides assumption-light
asymptotic coverage for the feature importance score of any model. Further, our
approach is fast as computations needed for inference come nearly for free as
part of the ensemble learning process. Finally, we also show that our same
procedure can be leveraged to provide valid confidence intervals for
predictions, hence providing fast, simultaneous quantification of the
uncertainty of both model predictions and interpretations. We validate our
intervals on a series of synthetic and real data examples, showing that our
approach detects the correct important features and exhibits many computational
and statistical advantages over existing methods. |
2,803 | Demystifying the Global Convergence Puzzle of Learning Over-parameterized ReLU Nets in Very High Dimensions | This theoretical paper is devoted to developing a rigorous theory for
demystifying the global convergence phenomenon in a challenging scenario:
learning over-parameterized Rectified Linear Unit (ReLU) nets for very high
dimensional dataset under very mild assumptions. A major ingredient of our
analysis is a fine-grained analysis of random activation matrices. The
essential virtue of dissecting activation matrices is that it bridges the
dynamics of optimization and angular distribution in high-dimensional data
space. This angle-based detailed analysis leads to asymptotic characterizations
of gradient norm and directional curvature of objective function at each
gradient descent iteration, revealing that the empirical loss function enjoys
nice geometrical properties in the overparameterized setting. Along the way, we
significantly improve existing theoretical bounds on both over-parameterization
condition and learning rate with very mild assumptions for learning very high
dimensional data. Moreover, we uncover the role of the geometrical and spectral
properties of the input data in determining desired over-parameterization size
and global convergence rate. All these clues allow us to discover a novel
geometric picture of nonconvex optimization in deep learning: angular
distribution in high-dimensional data space $\mapsto$ spectrums of
overparameterized activation matrices $\mapsto$ favorable geometrical
properties of empirical loss landscape $\mapsto$ global convergence phenomenon.
Furthremore, our theoretical results imply that gradient-based nonconvex
optimization algorithms have much stronger statistical guarantees with much
milder over-parameterization condition than exisiting theory states for
learning very high dimensional data, which is rarely explored so far. |
2,804 | Straggler-Resilient Personalized Federated Learning | Federated Learning is an emerging learning paradigm that allows training
models from samples distributed across a large network of clients while
respecting privacy and communication restrictions. Despite its success,
federated learning faces several challenges related to its decentralized
nature. In this work, we develop a novel algorithmic procedure with theoretical
speedup guarantees that simultaneously handles two of these hurdles, namely (i)
data heterogeneity, i.e., data distributions can vary substantially across
clients, and (ii) system heterogeneity, i.e., the computational power of the
clients could differ significantly. Our method relies on ideas from
representation learning theory to find a global common representation using all
clients' data and learn a user-specific set of parameters leading to a
personalized solution for each client. Furthermore, our method mitigates the
effects of stragglers by adaptively selecting clients based on their
computational characteristics and statistical significance, thus achieving, for
the first time, near optimal sample complexity and provable logarithmic
speedup. Experimental results support our theoretical findings showing the
superiority of our method over alternative personalized federated schemes in
system and data heterogeneous environments. |
2,805 | Deciding What to Model: Value-Equivalent Sampling for Reinforcement Learning | The quintessential model-based reinforcement-learning agent iteratively
refines its estimates or prior beliefs about the true underlying model of the
environment. Recent empirical successes in model-based reinforcement learning
with function approximation, however, eschew the true model in favor of a
surrogate that, while ignoring various facets of the environment, still
facilitates effective planning over behaviors. Recently formalized as the value
equivalence principle, this algorithmic technique is perhaps unavoidable as
real-world reinforcement learning demands consideration of a simple,
computationally-bounded agent interacting with an overwhelmingly complex
environment, whose underlying dynamics likely exceed the agent's capacity for
representation. In this work, we consider the scenario where agent limitations
may entirely preclude identifying an exactly value-equivalent model,
immediately giving rise to a trade-off between identifying a model that is
simple enough to learn while only incurring bounded sub-optimality. To address
this problem, we introduce an algorithm that, using rate-distortion theory,
iteratively computes an approximately-value-equivalent, lossy compression of
the environment which an agent may feasibly target in lieu of the true model.
We prove an information-theoretic, Bayesian regret bound for our algorithm that
holds for any finite-horizon, episodic sequential decision-making problem.
Crucially, our regret bound can be expressed in one of two possible forms,
providing a performance guarantee for finding either the simplest model that
achieves a desired sub-optimality gap or, alternatively, the best model given a
limit on agent capacity. |
2,806 | Learning Robust Representations Of Generative Models Using Set-Based Artificial Fingerprints | With recent progress in deep generative models, the problem of identifying
synthetic data and comparing their underlying generative processes has become
an imperative task for various reasons, including fighting visual
misinformation and source attribution. Existing methods often approximate the
distance between the models via their sample distributions. In this paper, we
approach the problem of fingerprinting generative models by learning
representations that encode the residual artifacts left by the generative
models as unique signals that identify the source models. We consider these
unique traces (a.k.a. "artificial fingerprints") as representations of
generative models, and demonstrate their usefulness in both the discriminative
task of source attribution and the unsupervised task of defining a similarity
between the underlying models. We first extend the existing studies on
fingerprints of GANs to four representative classes of generative models (VAEs,
Flows, GANs and score-based models), and demonstrate their existence and
attributability. We then improve the stability and attributability of the
fingerprints by proposing a new learning method based on set-encoding and
contrastive training. Our set-encoder, unlike existing methods that operate on
individual images, learns fingerprints from a \textit{set} of images. We
demonstrate improvements in the stability and attributability through
comparisons to state-of-the-art fingerprint methods and ablation studies.
Further, our method employs contrastive training to learn an implicit
similarity between models. We discover latent families of generative models
using this metric in a standard hierarchical clustering algorithm. |
2,807 | Active Bayesian Causal Inference | Causal discovery and causal reasoning are classically treated as separate and
consecutive tasks: one first infers the causal graph, and then uses it to
estimate causal effects of interventions. However, such a two-stage approach is
uneconomical, especially in terms of actively collected interventional data,
since the causal query of interest may not require a fully-specified causal
model. From a Bayesian perspective, it is also unnatural, since a causal query
(e.g., the causal graph or some causal effect) can be viewed as a latent
quantity subject to posterior inference -- other unobserved quantities that are
not of direct interest (e.g., the full causal model) ought to be marginalized
out in this process and contribute to our epistemic uncertainty. In this work,
we propose Active Bayesian Causal Inference (ABCI), a fully-Bayesian active
learning framework for integrated causal discovery and reasoning, which jointly
infers a posterior over causal models and queries of interest. In our approach
to ABCI, we focus on the class of causally-sufficient, nonlinear additive noise
models, which we model using Gaussian processes. We sequentially design
experiments that are maximally informative about our target causal query,
collect the corresponding interventional data, and update our beliefs to choose
the next experiment. Through simulations, we demonstrate that our approach is
more data-efficient than several baselines that only focus on learning the full
causal graph. This allows us to accurately learn downstream causal queries from
fewer samples while providing well-calibrated uncertainty estimates for the
quantities of interest. |
2,808 | AI in 6G: Energy-Efficient Distributed Machine Learning for Multilayer Heterogeneous Networks | Adept network management is key for supporting extremely heterogeneous
applications with stringent quality of service (QoS) requirements; this is more
so when envisioning the complex and ultra-dense 6G mobile heterogeneous network
(HetNet). From both the environmental and economical perspectives,
non-homogeneous QoS demands obstruct the minimization of the energy footprints
and operational costs of the envisioned robust networks. As such, network
intelligentization is expected to play an essential role in the realization of
such sophisticated aims. The fusion of artificial intelligence (AI) and mobile
networks will allow for the dynamic and automatic configuration of network
functionalities. Machine learning (ML), one of the backbones of AI, will be
instrumental in forecasting changes in network loads and resource utilization,
estimating channel conditions, optimizing network slicing, and enhancing
security and encryption. However, it is well known that ML tasks themselves
incur massive computational burdens and energy costs. To overcome such
obstacles, we propose a novel layer-based HetNet architecture which optimally
distributes tasks associated with different ML approaches across network layers
and entities; such a HetNet boasts multiple access schemes as well as
device-to-device (D2D) communications to enhance energy efficiency via
collaborative learning and communications. |
2,809 | Your Neighbors Are Communicating: Towards Powerful and Scalable Graph Neural Networks | Message passing graph neural networks (GNNs) are known to have their
expressiveness upper-bounded by 1-dimensional Weisfeiler-Lehman (1-WL)
algorithm. To achieve more powerful GNNs, existing attempts either require ad
hoc features, or involve operations that incur high time and space
complexities. In this work, we propose a general and provably powerful GNN
framework that preserves the scalability of message passing scheme. In
particular, we first propose to empower 1-WL for graph isomorphism test by
considering edges among neighbors, giving rise to NC-1-WL. The expressiveness
of NC-1-WL is shown to be strictly above 1-WL but below 3-WL theoretically.
Further, we propose the NC-GNN framework as a differentiable neural version of
NC-1-WL. Our simple implementation of NC-GNN is provably as powerful as
NC-1-WL. Experiments demonstrate that our NC-GNN achieves remarkable
performance on various benchmarks. |
2,810 | When Personalization Harms: Reconsidering the Use of Group Attributes in Prediction | The standard approach to personalization in machine learning consists of
training a model with group attributes like sex, age group, and blood type. In
this work, we show that this approach to personalization fails to improve
performance for all groups who provide personal data. We discuss how this
effect inflicts harm in applications where models assign predictions on the
basis of group membership. We propose collective preference guarantees to
ensure the fair use of group attributes in prediction. We characterize how
common approaches to personalization violate fair use due to failures in model
development and deployment. We conduct a comprehensive empirical study of
personalization in clinical prediction models. Our results highlight the
prevalence of fair use violations, demonstrate actionable interventions to
mitigate harm and underscore the need to measure the gains of personalization
for all groups who provide personal data. |
2,811 | Interpretable Models Capable of Handling Systematic Missingness in Imbalanced Classes and Heterogeneous Datasets | Application of interpretable machine learning techniques on medical datasets
facilitate early and fast diagnoses, along with getting deeper insight into the
data. Furthermore, the transparency of these models increase trust among
application domain experts. Medical datasets face common issues such as
heterogeneous measurements, imbalanced classes with limited sample size, and
missing data, which hinder the straightforward application of machine learning
techniques. In this paper we present a family of prototype-based (PB)
interpretable models which are capable of handling these issues. The models
introduced in this contribution show comparable or superior performance to
alternative techniques applicable in such situations. However, unlike ensemble
based models, which have to compromise on easy interpretation, the PB models
here do not. Moreover we propose a strategy of harnessing the power of
ensembles while maintaining the intrinsic interpretability of the PB models, by
averaging the model parameter manifolds. All the models were evaluated on a
synthetic (publicly available dataset) in addition to detailed analyses of two
real-world medical datasets (one publicly available). Results indicated that
the models and strategies we introduced addressed the challenges of real-world
medical data, while remaining computationally inexpensive and transparent, as
well as similar or superior in performance compared to their alternatives. |
2,812 | On the Generalization Power of the Overfitted Three-Layer Neural Tangent Kernel Model | In this paper, we study the generalization performance of overparameterized
3-layer NTK models. We show that, for a specific set of ground-truth functions
(which we refer to as the "learnable set"), the test error of the overfitted
3-layer NTK is upper bounded by an expression that decreases with the number of
neurons of the two hidden layers. Different from 2-layer NTK where there exists
only one hidden-layer, the 3-layer NTK involves interactions between two
hidden-layers. Our upper bound reveals that, between the two hidden-layers, the
test error descends faster with respect to the number of neurons in the second
hidden-layer (the one closer to the output) than with respect to that in the
first hidden-layer (the one closer to the input). We also show that the
learnable set of 3-layer NTK without bias is no smaller than that of 2-layer
NTK models with various choices of bias in the neurons. However, in terms of
the actual generalization performance, our results suggest that 3-layer NTK is
much less sensitive to the choices of bias than 2-layer NTK, especially when
the input dimension is large. |
2,813 | UAV-Aided Multi-Community Federated Learning | In this work, we investigate the problem of an online trajectory design for
an Unmanned Aerial Vehicle (UAV) in a Federated Learning (FL) setting where
several different communities exist, each defined by a unique task to be
learned. In this setting, spatially distributed devices belonging to each
community collaboratively contribute towards training their community model via
wireless links provided by the UAV. Accordingly, the UAV acts as a mobile
orchestrator coordinating the transmissions and the learning schedule among the
devices in each community, intending to accelerate the learning process of all
tasks. We propose a heuristic metric as a proxy for the training performance of
the different tasks. Capitalizing on this metric, a surrogate objective is
defined which enables us to jointly optimize the UAV trajectory and the
scheduling of the devices by employing convex optimization techniques and graph
theory. The simulations illustrate the out-performance of our solution when
compared to other handpicked static and mobile UAV deployment baselines. |
2,814 | Developing hierarchical anticipations via neural network-based event segmentation | Humans can make predictions on various time scales and hierarchical levels.
Thereby, the learning of event encodings seems to play a crucial role. In this
work we model the development of hierarchical predictions via autonomously
learned latent event codes. We present a hierarchical recurrent neural network
architecture, whose inductive learning biases foster the development of
sparsely changing latent state that compress sensorimotor sequences. A higher
level network learns to predict the situations in which the latent states tend
to change. Using a simulated robotic manipulator, we demonstrate that the
system (i) learns latent states that accurately reflect the event structure of
the data, (ii) develops meaningful temporal abstract predictions on the higher
level, and (iii) generates goal-anticipatory behavior similar to gaze behavior
found in eye-tracking studies with infants. The architecture offers a step
towards autonomous, self-motivated learning of compressed hierarchical
encodings of gathered experiences and the exploitation of these encodings for
the generation of highly versatile, adaptive behavior. |
2,815 | First-Order Algorithms for Min-Max Optimization in Geodesic Metric Spaces | From optimal transport to robust dimensionality reduction, a plethora of
machine learning applications can be cast into the min-max optimization
problems over Riemannian manifolds. Though many min-max algorithms have been
analyzed in the Euclidean setting, it has proved elusive to translate these
results to the Riemannian case. Zhang et al. [2022] have recently shown that
geodesic convex concave Riemannian problems always admit saddle-point
solutions. Inspired by this result, we study whether a performance gap between
Riemannian and optimal Euclidean space convex-concave algorithms is necessary.
We answer this question in the negative-we prove that the Riemannian corrected
extragradient (RCEG) method achieves last-iterate convergence at a linear rate
in the geodesically strongly-convex-concave case, matching the Euclidean
result. Our results also extend to the stochastic or non-smooth case where RCEG
and Riemanian gradient ascent descent (RGDA) achieve near-optimal convergence
rates up to factors depending on curvature of the manifold. |
2,816 | MetaNOR: A Meta-Learnt Nonlocal Operator Regression Approach for Metamaterial Modeling | We propose MetaNOR, a meta-learnt approach for transfer-learning operators
based on the nonlocal operator regression. The overall goal is to efficiently
provide surrogate models for new and unknown material-learning tasks with
different microstructures. The algorithm consists of two phases: (1) learning a
common nonlocal kernel representation from existing tasks; (2) transferring the
learned knowledge and rapidly learning surrogate operators for unseen tasks
with a different material, where only a few test samples are required. We apply
MetaNOR to model the wave propagation within 1D metamaterials, showing
substantial improvements on the sampling efficiency for new materials. |
2,817 | Beyond Value: CHECKLIST for Testing Inferences in Planning-Based RL | Reinforcement learning (RL) agents are commonly evaluated via their expected
value over a distribution of test scenarios. Unfortunately, this evaluation
approach provides limited evidence for post-deployment generalization beyond
the test distribution. In this paper, we address this limitation by extending
the recent CheckList testing methodology from natural language processing to
planning-based RL. Specifically, we consider testing RL agents that make
decisions via online tree search using a learned transition model and value
function. The key idea is to improve the assessment of future performance via a
CheckList approach for exploring and assessing the agent's inferences during
tree search. The approach provides the user with an interface and general
query-rule mechanism for identifying potential inference flaws and validating
expected inference invariances. We present a user study involving knowledgeable
AI researchers using the approach to evaluate an agent trained to play a
complex real-time strategy game. The results show the approach is effective in
allowing users to identify previously-unknown flaws in the agent's reasoning.
In addition, our analysis provides insight into how AI experts use this type of
testing approach, which may help improve future instantiations. |
2,818 | Interpolating Between Softmax Policy Gradient and Neural Replicator Dynamics with Capped Implicit Exploration | Neural replicator dynamics (NeuRD) is an alternative to the foundational
softmax policy gradient (SPG) algorithm motivated by online learning and
evolutionary game theory. The NeuRD expected update is designed to be nearly
identical to that of SPG, however, we show that the Monte Carlo updates differ
in a substantial way: the importance correction accounting for a sampled action
is nullified in the SPG update, but not in the NeuRD update. Naturally, this
causes the NeuRD update to have higher variance than its SPG counterpart.
Building on implicit exploration algorithms in the adversarial bandit setting,
we introduce capped implicit exploration (CIX) estimates that allow us to
construct NeuRD-CIX, which interpolates between this aspect of NeuRD and SPG.
We show how CIX estimates can be used in a black-box reduction to construct
bandit algorithms with regret bounds that hold with high probability and the
benefits this entails for NeuRD-CIX in sequential decision-making settings. Our
analysis reveals a bias--variance tradeoff between SPG and NeuRD, and shows how
theory predicts that NeuRD-CIX will perform well more consistently than NeuRD
while retaining NeuRD's advantages over SPG in non-stationary environments. |
2,819 | A Control Theoretic Framework for Adaptive Gradient Optimizers in Machine Learning | Adaptive gradient methods have become popular in optimizing deep neural
networks; recent examples include AdaGrad and Adam. Although Adam usually
converges faster, variations of Adam, for instance, the AdaBelief algorithm,
have been proposed to enhance Adam's poor generalization ability compared to
the classical stochastic gradient method. This paper develops a generic
framework for adaptive gradient methods that solve non-convex optimization
problems. We first model the adaptive gradient methods in a state-space
framework, which allows us to present simpler convergence proofs of adaptive
optimizers such as AdaGrad, Adam, and AdaBelief. We then utilize the transfer
function paradigm from classical control theory to propose a new variant of
Adam, coined AdamSSM. We add an appropriate pole-zero pair in the transfer
function from squared gradients to the second moment estimate. We prove the
convergence of the proposed AdamSSM algorithm. Applications on benchmark
machine learning tasks of image classification using CNN architectures and
language modeling using LSTM architecture demonstrate that the AdamSSM
algorithm improves the gap between generalization accuracy and faster
convergence than the recent adaptive gradient methods. |
2,820 | Zeroth-Order SciML: Non-intrusive Integration of Scientific Software with Deep Learning | Using deep learning (DL) to accelerate and/or improve scientific workflows
can yield discoveries that are otherwise impossible. Unfortunately, DL models
have yielded limited success in complex scientific domains due to large data
requirements. In this work, we propose to overcome this issue by integrating
the abundance of scientific knowledge sources (SKS) with the DL training
process. Existing knowledge integration approaches are limited to using
differentiable knowledge source to be compatible with first-order DL training
paradigm. In contrast, our proposed approach treats knowledge source as a
black-box in turn allowing to integrate virtually any knowledge source. To
enable an end-to-end training of SKS-coupled-DL, we propose to use zeroth-order
optimization (ZOO) based gradient-free training schemes, which is
non-intrusive, i.e., does not require making any changes to the SKS. We
evaluate the performance of our ZOO training scheme on two real-world material
science applications. We show that proposed scheme is able to effectively
integrate scientific knowledge with DL training and is able to outperform
purely data-driven model for data-limited scientific applications. We also
discuss some limitations of the proposed method and mention potentially
worthwhile future directions. |
2,821 | A Neural Network Approach for Homogenization of Multiscale Problems | We propose a neural network-based approach to the homogenization of
multiscale problems. The proposed method uses a derivative-free formulation of
a training loss, which incorporates Brownian walkers to find the macroscopic
description of a multiscale PDE solution. Compared with other network-based
approaches for multiscale problems, the proposed method is free from the design
of hand-crafted neural network architecture and the cell problem to calculate
the homogenization coefficient. The exploration neighborhood of the Brownian
walkers affects the overall learning trajectory. We determine the bounds of
micro- and macro-time steps that capture the local heterogeneous and global
homogeneous solution behaviors, respectively, through a neural network. The
bounds imply that the computational cost of the proposed method is independent
of the microscale periodic structure for the standard periodic problems. We
validate the efficiency and robustness of the proposed method through a suite
of linear and nonlinear multiscale problems with periodic and random field
coefficients. |
2,822 | Guided Deep Metric Learning | Deep Metric Learning (DML) methods have been proven relevant for visual
similarity learning. However, they sometimes lack generalization properties
because they are trained often using an inappropriate sample selection strategy
or due to the difficulty of the dataset caused by a distributional shift in the
data. These represent a significant drawback when attempting to learn the
underlying data manifold. Therefore, there is a pressing need to develop better
ways of obtaining generalization and representation of the underlying manifold.
In this paper, we propose a novel approach to DML that we call Guided Deep
Metric Learning, a novel architecture oriented to learning more compact
clusters, improving generalization under distributional shifts in DML. This
novel architecture consists of two independent models: A multi-branch master
model, inspired from a Few-Shot Learning (FSL) perspective, generates a reduced
hypothesis space based on prior knowledge from labeled data, which guides or
regularizes the decision boundary of a student model during training under an
offline knowledge distillation scheme. Experiments have shown that the proposed
method is capable of a better manifold generalization and representation to up
to 40% improvement (Recall@1, CIFAR10), using guidelines suggested by Musgrave
et al. to perform a more fair and realistic comparison, which is currently
absent in the literature |
2,823 | Between Rate-Distortion Theory & Value Equivalence in Model-Based Reinforcement Learning | The quintessential model-based reinforcement-learning agent iteratively
refines its estimates or prior beliefs about the true underlying model of the
environment. Recent empirical successes in model-based reinforcement learning
with function approximation, however, eschew the true model in favor of a
surrogate that, while ignoring various facets of the environment, still
facilitates effective planning over behaviors. Recently formalized as the value
equivalence principle, this algorithmic technique is perhaps unavoidable as
real-world reinforcement learning demands consideration of a simple,
computationally-bounded agent interacting with an overwhelmingly complex
environment. In this work, we entertain an extreme scenario wherein some
combination of immense environment complexity and limited agent capacity
entirely precludes identifying an exactly value-equivalent model. In light of
this, we embrace a notion of approximate value equivalence and introduce an
algorithm for incrementally synthesizing simple and useful approximations of
the environment from which an agent might still recover near-optimal behavior.
Crucially, we recognize the information-theoretic nature of this lossy
environment compression problem and use the appropriate tools of
rate-distortion theory to make mathematically precise how value equivalence can
lend tractability to otherwise intractable sequential decision-making problems. |
2,824 | Is $L^2$ Physics-Informed Loss Always Suitable for Training Physics-Informed Neural Network? | The Physics-Informed Neural Network (PINN) approach is a new and promising
way to solve partial differential equations using deep learning. The $L^2$
Physics-Informed Loss is the de-facto standard in training Physics-Informed
Neural Networks. In this paper, we challenge this common practice by
investigating the relationship between the loss function and the approximation
quality of the learned solution. In particular, we leverage the concept of
stability in the literature of partial differential equation to study the
asymptotic behavior of the learned solution as the loss approaches zero. With
this concept, we study an important class of high-dimensional non-linear PDEs
in optimal control, the Hamilton-Jacobi-Bellman(HJB) Equation, and prove that
for general $L^p$ Physics-Informed Loss, a wide class of HJB equation is stable
only if $p$ is sufficiently large. Therefore, the commonly used $L^2$ loss is
not suitable for training PINN on those equations, while $L^{\infty}$ loss is a
better choice. Based on the theoretical insight, we develop a novel PINN
training algorithm to minimize the $L^{\infty}$ loss for HJB equations which is
in a similar spirit to adversarial training. The effectiveness of the proposed
algorithm is empirically demonstrated through experiments. |
2,825 | Causal Discovery in Heterogeneous Environments Under the Sparse Mechanism Shift Hypothesis | Machine learning approaches commonly rely on the assumption of independent
and identically distributed (i.i.d.) data. In reality, however, this assumption
is almost always violated due to distribution shifts between environments.
Although valuable learning signals can be provided by heterogeneous data from
changing distributions, it is also known that learning under arbitrary
(adversarial) changes is impossible. Causality provides a useful framework for
modeling distribution shifts, since causal models encode both observational and
interventional distributions. In this work, we explore the sparse mechanism
shift hypothesis, which posits that distribution shifts occur due to a small
number of changing causal conditionals. Motivated by this idea, we apply it to
learning causal structure from heterogeneous environments, where i.i.d. data
only allows for learning an equivalence class of graphs without restrictive
assumptions. We propose the Mechanism Shift Score (MSS), a score-based approach
amenable to various empirical estimators, which provably identifies the entire
causal structure with high probability if the sparse mechanism shift hypothesis
holds. Empirically, we verify behavior predicted by the theory and compare
multiple estimators and score functions to identify the best approaches in
practice. Compared to other methods, we show how MSS bridges a gap by both
being nonparametric as well as explicitly leveraging sparse changes. |
2,826 | Combinatorial optimization for low bit-width neural networks | Low-bit width neural networks have been extensively explored for deployment
on edge devices to reduce computational resources. Existing approaches have
focused on gradient-based optimization in a two-stage train-and-compress
setting or as a combined optimization where gradients are quantized during
training. Such schemes require high-performance hardware during the training
phase and usually store an equivalent number of full-precision weights apart
from the quantized weights. In this paper, we explore methods of direct
combinatorial optimization in the problem of risk minimization with binary
weights, which can be made equivalent to a non-monotone submodular maximization
under certain conditions. We employ an approximation algorithm for the cases
with single and multilayer neural networks. For linear models, it has
$\mathcal{O}(nd)$ time complexity where $n$ is the sample size and $d$ is the
data dimension. We show that a combination of greedy coordinate descent and
this novel approach can attain competitive accuracy on binary classification
tasks. |
2,827 | CVNets: High Performance Library for Computer Vision | We introduce CVNets, a high-performance open-source library for training deep
neural networks for visual recognition tasks, including classification,
detection, and segmentation. CVNets supports image and video understanding
tools, including data loading, data transformations, novel data sampling
methods, and implementations of several standard networks with similar or
better performance than previous studies.
Our source code is available at: \url{https://github.com/apple/ml-cvnets}. |
2,828 | Hybrid Value Estimation for Off-policy Evaluation and Offline Reinforcement Learning | Value function estimation is an indispensable subroutine in reinforcement
learning, which becomes more challenging in the offline setting. In this paper,
we propose Hybrid Value Estimation (HVE) to reduce value estimation error,
which trades off bias and variance by balancing between the value estimation
from offline data and the learned model. Theoretical analysis discloses that
HVE enjoys a better error bound than the direct methods. HVE can be leveraged
in both off-policy evaluation and offline reinforcement learning settings. We,
therefore, provide two concrete algorithms Off-policy HVE (OPHVE) and
Model-based Offline HVE (MOHVE), respectively. Empirical evaluations on MuJoCo
tasks corroborate the theoretical claim. OPHVE outperforms other off-policy
evaluation methods in all three metrics measuring the estimation effectiveness,
while MOHVE achieves better or comparable performance with state-of-the-art
offline reinforcement learning algorithms. We hope that HVE could shed some
light on further research on reinforcement learning from fixed data. |
2,829 | MSR: Making Self-supervised learning Robust to Aggressive Augmentations | Most recent self-supervised learning methods learn visual representation by
contrasting different augmented views of images. Compared with supervised
learning, more aggressive augmentations have been introduced to further improve
the diversity of training pairs. However, aggressive augmentations may distort
images' structures leading to a severe semantic shift problem that augmented
views of the same image may not share the same semantics, thus degrading the
transfer performance. To address this problem, we propose a new SSL paradigm,
which counteracts the impact of semantic shift by balancing the role of weak
and aggressively augmented pairs. Specifically, semantically inconsistent pairs
are of minority and we treat them as noisy pairs. Note that deep neural
networks (DNNs) have a crucial memorization effect that DNNs tend to first
memorize clean (majority) examples before overfitting to noisy (minority)
examples. Therefore, we set a relatively large weight for aggressively
augmented data pairs at the early learning stage. With the training going on,
the model begins to overfit noisy pairs. Accordingly, we gradually reduce the
weights of aggressively augmented pairs. In doing so, our method can better
embrace the aggressive augmentations and neutralize the semantic shift problem.
Experiments show that our model achieves 73.1% top-1 accuracy on ImageNet-1K
with ResNet-50 for 200 epochs, which is a 2.5% improvement over BYOL. Moreover,
experiments also demonstrate that the learned representations can transfer well
for various downstream tasks. |
2,830 | Combinatorial Causal Bandits | In combinatorial causal bandits (CCB), the learning agent chooses at most $K$
variables in each round to intervene, collects feedback from the observed
variables, with the goal of minimizing expected regret on the target variable
$Y$. Different from all prior studies on causal bandits, CCB needs to deal with
exponentially large action space. We study under the context of binary
generalized linear models (BGLMs) with a succinct parametric representation of
the causal models. We present the algorithm BGLM-OFU for Markovian BGLMs (i.e.
no hidden variables) based on the maximum likelihood estimation method, and
show that it achieves $O(\sqrt{T}\log T)$ regret, where $T$ is the time
horizon. For the special case of linear models with hidden variables, we apply
causal inference techniques such as the do-calculus to convert the original
model into a Markovian model, and then show that our BGLM-OFU algorithm and
another algorithm based on the linear regression both solve such linear models
with hidden variables. Our novelty includes (a) considering the combinatorial
intervention action space, (b) considering general causal models including ones
with hidden variables, (c) integrating and adapting techniques from diverse
studies such as generalized linear bandits and online influence maximization,
and (d) not relying on unrealistic assumptions such as knowing the joint
distribution of the parents of $Y$ under all interventions used in some prior
studies. |
2,831 | Rethinking the Openness of CLIP | Contrastive Language-Image Pre-training (CLIP) has demonstrated great
potential in realizing open-vocabulary image classification in a matching
style, because of its holistic use of natural language supervision that covers
unconstrained real-world visual concepts. However, it is, in turn, also
difficult to evaluate and analyze the openness of CLIP-like models, since they
are in theory open to any vocabulary but the actual accuracy varies. To address
the insufficiency of conventional studies on openness, we resort to an
incremental view and define the extensibility, which essentially approximates
the model's ability to deal with new visual concepts, by evaluating openness
through vocabulary expansions. Our evaluation based on extensibility shows that
CLIP-like models are hardly truly open and their performances degrade as the
vocabulary expands to different degrees. Further analysis reveals that the
over-estimation of openness is not because CLIP-like models fail to capture the
general similarity of image and text features of novel visual concepts, but
because of the confusion among competing text features, that is, they are not
stable with respect to the vocabulary. In light of this, we propose to improve
the openness of CLIP from the perspective of feature space by enforcing the
distinguishability of text features. Our method retrieves relevant texts from
the pre-training corpus to enhance prompts for inference, which boosts the
extensibility and stability of CLIP even without fine-tuning. |
2,832 | Geodesic Properties of a Generalized Wasserstein Embedding for Time Series Analysis | Transport-based metrics and related embeddings (transforms) have recently
been used to model signal classes where nonlinear structures or variations are
present. In this paper, we study the geodesic properties of time series data
with a generalized Wasserstein metric and the geometry related to their signed
cumulative distribution transforms in the embedding space. Moreover, we show
how understanding such geometric characteristics can provide added
interpretability to certain time series classifiers, and be an inspiration for
more robust classifiers. |
2,833 | Modelling and Mining of Patient Pathways: A Scoping Review | The sequence of visits and procedures performed by the patient in the health
system, also known as the patient's pathway or trajectory, can reveal important
information about the clinical treatment adopted and the health service
provided. The rise of electronic health data availability made it possible to
assess the pathways of a large number of patients. Nevertheless, some
challenges also arose concerning how to synthesize these pathways and how to
mine them from the data, fostering a new field of research. The objective of
this review is to survey this new field of research, highlighting
representation models, mining techniques, methods of analysis, and examples of
case studies. |
2,834 | Future Artificial Intelligence tools and perspectives in medicine | Purpose of review: Artificial intelligence (AI) has become popular in medical
applications, specifically as a clinical support tool for computer-aided
diagnosis. These tools are typically employed on medical data (i.e., image,
molecular data, clinical variables, etc.) and used the statistical and machine
learning methods to measure the model performance. In this review, we
summarized and discussed the most recent radiomic pipeline used for clinical
analysis. Recent findings:Currently, limited management of cancers benefits
from artificial intelligence, mostly related to a computer-aided diagnosis that
avoids a biopsy analysis that presents additional risks and costs. Most AI
tools are based on imaging features, known as radiomic analysis that can be
refined into predictive models in non-invasively acquired imaging data. This
review explores the progress of AI-based radiomic tools for clinical
applications with a brief description of necessary technical steps. Explaining
new radiomic approaches based on deep learning techniques will explain how the
new radiomic models (deep radiomic analysis) can benefit from deep
convolutional neural networks and be applied on limited data sets. Summary: To
consider the radiomic algorithms, further investigations are recommended to
involve deep learning in radiomic models with additional validation steps on
various cancer types. |
2,835 | Formal Specifications from Natural Language | We study the ability of language models to translate natural language into
formal specifications with complex semantics. In particular, we fine-tune
off-the-shelf language models on three datasets consisting of structured
English sentences and their corresponding formal representation: 1) First-order
logic (FOL), commonly used in software verification and theorem proving; 2)
linear-time temporal logic (LTL), which forms the basis for industrial hardware
specification languages; and 3) regular expressions (regex), frequently used in
programming and search. Our experiments show that, in these diverse domains,
the language models achieve competitive performance to the respective
state-of-the-art with the benefits of being easy to access, cheap to fine-tune,
and without a particular need for domain-specific reasoning. Additionally, we
show that the language models have a unique selling point: they benefit from
their generalization capabilities from pre-trained knowledge on natural
language, e.g., to generalize to unseen variable names. |
2,836 | C$^3$Fusion: Consistent Contrastive Colon Fusion, Towards Deep SLAM in Colonoscopy | 3D colon reconstruction from Optical Colonoscopy (OC) to detect non-examined
surfaces remains an unsolved problem. The challenges arise from the nature of
optical colonoscopy data, characterized by highly reflective low-texture
surfaces, drastic illumination changes and frequent tracking loss. Recent
methods demonstrate compelling results, but suffer from: (1) frangible
frame-to-frame (or frame-to-model) pose estimation resulting in many tracking
failures; or (2) rely on point-based representations at the cost of scan
quality. In this paper, we propose a novel reconstruction framework that
addresses these issues end to end, which result in both quantitatively and
qualitatively accurate and robust 3D colon reconstruction. Our SLAM approach,
which employs correspondences based on contrastive deep features, and deep
consistent depth maps, estimates globally optimized poses, is able to recover
from frequent tracking failures, and estimates a global consistent 3D model;
all within a single framework. We perform an extensive experimental evaluation
on multiple synthetic and real colonoscopy videos, showing high-quality results
and comparisons against relevant baselines. |
2,837 | Comparing Performance of Different Linguistically-Backed Word Embeddings for Cyberbullying Detection | In most cases, word embeddings are learned only from raw tokens or in some
cases, lemmas. This includes pre-trained language models like BERT. To
investigate on the potential of capturing deeper relations between lexical
items and structures and to filter out redundant information, we propose to
preserve the morphological, syntactic and other types of linguistic information
by combining them with the raw tokens or lemmas. This means, for example,
including parts-of-speech or dependency information within the used lexical
features. The word embeddings can then be trained on the combinations instead
of just raw tokens. It is also possible to later apply this method to the
pre-training of huge language models and possibly enhance their performance.
This would aid in tackling problems which are more sophisticated from the point
of view of linguistic representation, such as detection of cyberbullying. |
2,838 | Exploring the Potential of Feature Density in Estimating Machine Learning Classifier Performance with Application to Cyberbullying Detection | In this research. we analyze the potential of Feature Density (HD) as a way
to comparatively estimate machine learning (ML) classifier performance prior to
training. The goal of the study is to aid in solving the problem of
resource-intensive training of ML models which is becoming a serious issue due
to continuously increasing dataset sizes and the ever rising popularity of Deep
Neural Networks (DNN). The issue of constantly increasing demands for more
powerful computational resources is also affecting the environment, as training
large-scale ML models are causing alarmingly-growing amounts of CO2, emissions.
Our approach 1s to optimize the resource-intensive training of ML models for
Natural Language Processing to reduce the number of required experiments
iterations. We expand on previous attempts on improving classifier training
efficiency with FD while also providing an insight to the effectiveness of
various linguistically-backed feature preprocessing methods for dialog
classification, specifically cyberbullying detection. |
2,839 | Robust Meta-learning with Sampling Noise and Label Noise via Eigen-Reptile | Recent years have seen a surge of interest in meta-learning techniques for
tackling the few-shot learning (FSL) problem. However, the meta-learner is
prone to overfitting since there are only a few available samples, which can be
identified as sampling noise on a clean dataset. Moreover, when handling the
data with noisy labels, the meta-learner could be extremely sensitive to label
noise on a corrupted dataset. To address these two challenges, we present
Eigen-Reptile (ER) that updates the meta-parameters with the main direction of
historical task-specific parameters to alleviate sampling and label noise.
Specifically, the main direction is computed in a fast way, where the scale of
the calculated matrix is related to the number of gradient steps instead of the
number of parameters. Furthermore, to obtain a more accurate main direction for
Eigen-Reptile in the presence of many noisy labels, we further propose
Introspective Self-paced Learning (ISPL). We have theoretically and
experimentally demonstrated the soundness and effectiveness of the proposed
Eigen-Reptile and ISPL. Particularly, our experiments on different tasks show
that the proposed method is able to outperform or achieve highly competitive
performance compared with other gradient-based methods with or without noisy
labels. The code and data for the proposed method are provided for research
purposes https://github.com/Anfeather/Eigen-Reptile. |
2,840 | Learning Generative Factors of Neuroimaging Data with Variational auto-encoders | Neuroimaging techniques produce high-dimensional, stochastic data from which
it might be challenging to extract high-level knowledge about the phenomena of
interest. We address this challenge by applying the generative modelling
framework to 1) classify multiple pathologies and 2) recover the neurological
mechanisms of those pathologies in a data-driven manner. Our framework learns
generative factors of data related to pathologies. We provide an algorithm to
decode those factors further and observe how different pathologies affect
observed data. We illustrate the applicability of the proposed approach to
identifying schizophrenia, either followed or not by auditory verbal
hallucinations. We further demonstrate the ability of the framework to learn
disease-related mechanisms consistent with current domain knowledge. We also
compare the proposed framework with several benchmark approaches and indicate
its classification performance and interpretability advantages. |
2,841 | Stochastic Multiple Target Sampling Gradient Descent | Sampling from an unnormalized target distribution is an essential problem
with many applications in probabilistic inference. Stein Variational Gradient
Descent (SVGD) has been shown to be a powerful method that iteratively updates
a set of particles to approximate the distribution of interest. Furthermore,
when analysing its asymptotic properties, SVGD reduces exactly to a
single-objective optimization problem and can be viewed as a probabilistic
version of this single-objective optimization problem. A natural question then
arises: "Can we derive a probabilistic version of the multi-objective
optimization?". To answer this question, we propose Stochastic Multiple Target
Sampling Gradient Descent (MT-SGD), enabling us to sample from multiple
unnormalized target distributions. Specifically, our MT-SGD conducts a flow of
intermediate distributions gradually orienting to multiple target
distributions, which allows the sampled particles to move to the joint
high-likelihood region of the target distributions. Interestingly, the
asymptotic analysis shows that our approach reduces exactly to the
multiple-gradient descent algorithm for multi-objective optimization, as
expected. Finally, we conduct comprehensive experiments to demonstrate the
merit of our approach to multi-task learning. |
2,842 | Investigating Brain Connectivity with Graph Neural Networks and GNNExplainer | Functional connectivity plays an essential role in modern neuroscience. The
modality sheds light on the brain's functional and structural aspects,
including mechanisms behind multiple pathologies. One such pathology is
schizophrenia which is often followed by auditory verbal hallucinations. The
latter is commonly studied by observing functional connectivity during speech
processing. In this work, we have made a step toward an in-depth examination of
functional connectivity during a dichotic listening task via deep learning for
three groups of people: schizophrenia patients with and without auditory verbal
hallucinations and healthy controls. We propose a graph neural network-based
framework within which we represent EEG data as signals in the graph domain.
The framework allows one to 1) predict a brain mental disorder based on EEG
recording, 2) differentiate the listening state from the resting state for each
group and 3) recognize characteristic task-depending connectivity. Experimental
results show that the proposed model can differentiate between the above groups
with state-of-the-art performance. Besides, it provides a researcher with
meaningful information regarding each group's functional connectivity, which we
validated on the current domain knowledge. |
2,843 | Variational Monte Carlo Approach to Partial Differential Equations with Neural Networks | The accurate numerical solution of partial differential equations is a
central task in numerical analysis allowing to model a wide range of natural
phenomena by employing specialized solvers depending on the scenario of
application. Here, we develop a variational approach for solving partial
differential equations governing the evolution of high dimensional probability
distributions. Our approach naturally works on the unbounded continuous domain
and encodes the full probability density function through its variational
parameters, which are adapted dynamically during the evolution to optimally
reflect the dynamics of the density. For the considered benchmark cases we
observe excellent agreement with numerical solutions as well as analytical
solutions in regimes inaccessible to traditional computational approaches. |
2,844 | Classification at the Accuracy Limit -- Facing the Problem of Data Ambiguity | Data classification, the process of analyzing data and organizing it into
categories, is a fundamental computing problem of natural and artificial
information processing systems. Ideally, the performance of classifier models
would be evaluated using unambiguous data sets, where the 'correct' assignment
of category labels to the input data vectors is unequivocal. In real-world
problems, however, a significant fraction of actually occurring data vectors
will be located in a boundary zone between or outside of all categories, so
that perfect classification cannot even in principle be achieved. We derive the
theoretical limit for classification accuracy that arises from the overlap of
data categories. By using a surrogate data generation model with adjustable
statistical properties, we show that sufficiently powerful classifiers based on
completely different principles, such as perceptrons and Bayesian models, all
perform at this universal accuracy limit. Remarkably, the accuracy limit is not
affected by applying non-linear transformations to the data, even if these
transformations are non-reversible and drastically reduce the information
content of the input data. We compare emerging data embeddings produced by
supervised and unsupervised training, using MNIST and human EEG recordings
during sleep. We find that categories are not only well separated in the final
layers of classifiers trained with back-propagation, but to a smaller degree
also after unsupervised dimensionality reduction. This suggests that
human-defined categories, such as hand-written digits or sleep stages, can
indeed be considered as 'natural kinds'. |
2,845 | Neural Lyapunov Control of Unknown Nonlinear Systems with Stability Guarantees | Learning for control of dynamical systems with formal guarantees remains a
challenging task. This paper proposes a learning framework to simultaneously
stabilize an unknown nonlinear system with a neural controller and learn a
neural Lyapunov function to certify a region of attraction (ROA) for the
closed-loop system. The algorithmic structure consists of two neural networks
and a satisfiability modulo theories (SMT) solver. The first neural network is
responsible for learning the unknown dynamics. The second neural network aims
to identify a valid Lyapunov function and a provably stabilizing nonlinear
controller. The SMT solver then verifies that the candidate Lyapunov function
indeed satisfies the Lyapunov conditions. We provide theoretical guarantees of
the proposed learning framework in terms of the closed-loop stability for the
unknown nonlinear system. We illustrate the effectiveness of the approach with
a set of numerical experiments. |
2,846 | Toward Learning Robust and Invariant Representations with Alignment Regularization and Data Augmentation | Data augmentation has been proven to be an effective technique for developing
machine learning models that are robust to known classes of distributional
shifts (e.g., rotations of images), and alignment regularization is a technique
often used together with data augmentation to further help the model learn
representations invariant to the shifts used to augment the data. In this
paper, motivated by a proliferation of options of alignment regularizations, we
seek to evaluate the performances of several popular design choices along the
dimensions of robustness and invariance, for which we introduce a new test
procedure. Our synthetic experiment results speak to the benefits of squared l2
norm regularization. Further, we also formally analyze the behavior of
alignment regularization to complement our empirical study under assumptions we
consider realistic. Finally, we test this simple technique we identify
(worst-case data augmentation with squared l2 norm alignment regularization)
and show that the benefits of this method outrun those of the specially
designed methods. We also release a software package in both TensorFlow and
PyTorch for users to use the method with a couple of lines at
https://github.com/jyanln/AlignReg. |
2,847 | Hybrid Architectures for Distributed Machine Learning in Heterogeneous Wireless Networks | The ever-growing data privacy concerns have transformed machine learning (ML)
architectures from centralized to distributed, leading to federated learning
(FL) and split learning (SL) as the two most popular privacy-preserving ML
paradigms. However, implementing either conventional FL or SL alone with
diverse network conditions (e.g., device-to-device (D2D) and cellular
communications) and heterogeneous clients (e.g., heterogeneous
computation/communication/energy capabilities) may face significant challenges,
particularly poor architecture scalability and long training time. To this end,
this article proposes two novel hybrid distributed ML architectures, namely,
hybrid split FL (HSFL) and hybrid federated SL (HFSL), by combining the
advantages of both FL and SL in D2D-enabled heterogeneous wireless networks.
Specifically, the performance comparison and advantages of HSFL and HFSL are
analyzed generally. Promising open research directions are presented to offer
commendable reference for future research. Finally, primary simulations are
conducted upon considering three datasets under non-independent and identically
distributed settings, to verify the feasibility of our proposed architectures,
which can significantly reduce communication/computation cost and training
time, as compared with conventional FL and SL. |
2,848 | Soft Adversarial Training Can Retain Natural Accuracy | Adversarial training for neural networks has been in the limelight in recent
years. The advancement in neural network architectures over the last decade has
led to significant improvement in their performance. It sparked an interest in
their deployment for real-time applications. This process initiated the need to
understand the vulnerability of these models to adversarial attacks. It is
instrumental in designing models that are robust against adversaries. Recent
works have proposed novel techniques to counter the adversaries, most often
sacrificing natural accuracy. Most suggest training with an adversarial version
of the inputs, constantly moving away from the original distribution. The focus
of our work is to use abstract certification to extract a subset of inputs for
(hence we call it 'soft') adversarial training. We propose a training framework
that can retain natural accuracy without sacrificing robustness in a
constrained setting. Our framework specifically targets moderately critical
applications which require a reasonable balance between robustness and
accuracy. The results testify to the idea of soft adversarial training for the
defense against adversarial attacks. At last, we propose the scope of future
work for further improvement of this framework. |
2,849 | Estimating counterfactual treatment outcomes over time in complex multi-agent scenarios | Evaluation of intervention in a multi-agent system, e.g., when humans should
intervene in autonomous driving systems and when a player should pass to
teammates for a good shot, is challenging in various engineering and scientific
fields. Estimating the individual treatment effect (ITE) using counterfactual
long-term prediction is practical to evaluate such interventions. However, most
of the conventional frameworks did not consider the time-varying complex
structure of multi-agent relationships and covariate counterfactual prediction.
This may sometimes lead to erroneous assessments of ITE and interpretation
problems. Here we propose an interpretable, counterfactual recurrent network in
multi-agent systems to estimate the effect of the intervention. Our model
leverages graph variational recurrent neural networks and theory-based
computation with domain knowledge for the ITE estimation framework based on
long-term prediction of multi-agent covariates and outcomes, which can confirm
under the circumstances under which the intervention is effective. On simulated
models of an automated vehicle and biological agents with time-varying
confounders, we show that our methods achieved lower estimation errors in
counterfactual covariates and the most effective treatment timing than the
baselines. Furthermore, using real basketball data, our methods performed
realistic counterfactual predictions and evaluated the counterfactual passes in
shot scenarios. |
2,850 | Evaluation of creating scoring opportunities for teammates in soccer via trajectory prediction | Evaluating the individual movements for teammates in soccer players is
crucial for assessing teamwork, scouting, and fan engagement. It has been said
that players in a 90-min game do not have the ball for about 87 minutes on
average. However, it has remained difficult to evaluate an attacking player
without receiving the ball, and to reveal how movement contributes to the
creation of scoring opportunities for teammates. In this paper, we evaluate
players who create off-ball scoring opportunities by comparing actual movements
with the reference movements generated via trajectory prediction. First, we
predict the trajectories of players using a graph variational recurrent neural
network that can accurately model the relationship between players and predict
the long-term trajectory. Next, based on the difference in the modified
off-ball evaluation index between the actual and the predicted trajectory as a
reference, we evaluate how the actual movement contributes to scoring
opportunity compared to the predicted movement. For verification, we examined
the relationship with the annual salary, the goals, and the rating in the game
by experts for all games of a team in a professional soccer league in a year.
The results show that the annual salary and the proposed indicator correlated
significantly, which could not be explained by the existing indicators and
goals. Our results suggest the effectiveness of the proposed method as an
indicator for a player without the ball to create a scoring chance for
teammates. |
2,851 | Saliency Attack: Towards Imperceptible Black-box Adversarial Attack | Deep neural networks are vulnerable to adversarial examples, even in the
black-box setting where the attacker is only accessible to the model output.
Recent studies have devised effective black-box attacks with high query
efficiency. However, such performance is often accompanied by compromises in
attack imperceptibility, hindering the practical use of these approaches. In
this paper, we propose to restrict the perturbations to a small salient region
to generate adversarial examples that can hardly be perceived. This approach is
readily compatible with many existing black-box attacks and can significantly
improve their imperceptibility with little degradation in attack success rate.
Further, we propose the Saliency Attack, a new black-box attack aiming to
refine the perturbations in the salient region to achieve even better
imperceptibility. Extensive experiments show that compared to the
state-of-the-art black-box attacks, our approach achieves much better
imperceptibility scores, including most apparent distortion (MAD), $L_0$ and
$L_2$ distances, and also obtains significantly higher success rates judged by
a human-like threshold on MAD. Importantly, the perturbations generated by our
approach are interpretable to some extent. Finally, it is also demonstrated to
be robust to different detection-based defenses. |
2,852 | Adaptive Tree Backup Algorithms for Temporal-Difference Reinforcement Learning | Q($\sigma$) is a recently proposed temporal-difference learning method that
interpolates between learning from expected backups and sampled backups. It has
been shown that intermediate values for the interpolation parameter $\sigma \in
[0,1]$ perform better in practice, and therefore it is commonly believed that
$\sigma$ functions as a bias-variance trade-off parameter to achieve these
improvements. In our work, we disprove this notion, showing that the choice of
$\sigma=0$ minimizes variance without increasing bias. This indicates that
$\sigma$ must have some other effect on learning that is not fully understood.
As an alternative, we hypothesize the existence of a new trade-off: larger
$\sigma$-values help overcome poor initializations of the value function, at
the expense of higher statistical variance. To automatically balance these
considerations, we propose Adaptive Tree Backup (ATB) methods, whose weighted
backups evolve as the agent gains experience. Our experiments demonstrate that
adaptive strategies can be more effective than relying on fixed or
time-annealed $\sigma$-values. |
2,853 | Initial Study into Application of Feature Density and Linguistically-backed Embedding to Improve Machine Learning-based Cyberbullying Detection | In this research, we study the change in the performance of machine learning
(ML) classifiers when various linguistic preprocessing methods of a dataset
were used, with the specific focus on linguistically-backed embeddings in
Convolutional Neural Networks (CNN). Moreover, we study the concept of Feature
Density and confirm its potential to comparatively predict the performance of
ML classifiers, including CNN. The research was conducted on a Formspring
dataset provided in a Kaggle competition on automatic cyberbullying detection.
The dataset was re-annotated by objective experts (psychologists), as the
importance of professional annotation in cyberbullying research has been
indicated multiple times. The study confirmed the effectiveness of Neural
Networks in cyberbullying detection and the correlation between classifier
performance and Feature Density while also proposing a new approach of training
various linguistically-backed embeddings for Convolutional Neural Networks. |
2,854 | Reward Poisoning Attacks on Offline Multi-Agent Reinforcement Learning | We expose the danger of reward poisoning in offline multi-agent reinforcement
learning (MARL), whereby an attacker can modify the reward vectors to different
learners in an offline data set while incurring a poisoning cost. Based on the
poisoned data set, all rational learners using some confidence-bound-based MARL
algorithm will infer that a target policy - chosen by the attacker and not
necessarily a solution concept originally - is the Markov perfect dominant
strategy equilibrium for the underlying Markov Game, hence they will adopt this
potentially damaging target policy in the future. We characterize the exact
conditions under which the attacker can install a target policy. We further
show how the attacker can formulate a linear program to minimize its poisoning
cost. Our work shows the need for robust MARL against adversarial attacks. |
2,855 | Learning in Congestion Games with Bandit Feedback | Learning Nash equilibria is a central problem in multi-agent systems. In this
paper, we investigate congestion games, a class of games with benign
theoretical structure and broad real-world applications. We first propose a
centralized algorithm based on the optimism in the face of uncertainty
principle for congestion games with (semi-)bandit feedback, and obtain
finite-sample guarantees. Then we propose a decentralized algorithm via a novel
combination of the Frank-Wolfe method and G-optimal design. By exploiting the
structure of the congestion game, we show the sample complexity of both
algorithms depends only polynomially on the number of players and the number of
facilities, but not the size of the action set, which can be exponentially
large in terms of the number of facilities. We further define a new problem
class, Markov congestion games, which allows us to model the non-stationarity
in congestion games. We propose a centralized algorithm for Markov congestion
games, whose sample complexity again has only polynomial dependence on all
relevant problem parameters, but not the size of the action set. |
2,856 | An Unpooling Layer for Graph Generation | We propose a novel and trainable graph unpooling layer for effective graph
generation. Given a graph with features, the unpooling layer enlarges this
graph and learns its desired new structure and features. Since this unpooling
layer is trainable, it can be applied to graph generation either in the decoder
of a variational autoencoder or in the generator of a generative adversarial
network (GAN). We prove that the unpooled graph remains connected and any
connected graph can be sequentially unpooled from a 3-nodes graph. We apply the
unpooling layer within the GAN generator. Since the most studied instance of
graph generation is molecular generation, we test our ideas in this context.
Using the QM9 and ZINC datasets, we demonstrate the improvement obtained by
using the unpooling layer instead of an adjacency-matrix-based approach. |
2,857 | Estimating the Effect of Team Hitting Strategies Using Counterfactual Virtual Simulation in Baseball | In baseball, every play on the field is quantitatively evaluated and has an
effect on individual and team strategies. The weighted on base average (wOBA)
is well known as a measure of an batter's hitting contribution. However, this
measure ignores the game situation, such as the runners on base, which coaches
and batters are known to consider when employing multiple hitting strategies,
yet, the effectiveness of these strategies is unknown. This is probably because
(1) we cannot obtain the batter's strategy and (2) it is difficult to estimate
the effect of the strategies. Here, we propose a new method for estimating the
effect using counterfactual batting simulation. To this end, we propose a deep
learning model that transforms batting ability when batting strategy is
changed. This method can estimate the effects of various strategies, which has
been traditionally difficult with actual game data. We found that, when the
switching cost of batting strategies can be ignored, the use of different
strategies increased runs. When the switching cost is considered, the
conditions for increasing runs were limited. Our validation results suggest
that our simulation could clarify the effect of using multiple batting
strategies. |
2,858 | Model-Informed Generative Adversarial Network (MI-GAN) for Learning Optimal Power Flow | The optimal power flow (OPF) problem, as a critical component of power system
operations, becomes increasingly difficult to solve due to the variability,
intermittency, and unpredictability of renewable energy brought to the power
system. Although traditional optimization techniques, such as stochastic and
robust optimization approaches, could be used to address the OPF problem in the
face of renewable energy uncertainty, their effectiveness in dealing with
large-scale problems remains limited. As a result, deep learning techniques,
such as neural networks, have recently been developed to improve computational
efficiency in solving large-scale OPF problems. However, the feasibility and
optimality of the solution may not be guaranteed. In this paper, we propose an
optimization model-informed generative adversarial network (MI-GAN) framework
to solve OPF under uncertainty. The main contributions are summarized into
three aspects: (1) to ensure feasibility and improve optimality of generated
solutions, three important layers are proposed: feasibility filter layer,
comparison layer, and gradient-guided layer; (2) in the GAN-based framework, an
efficient model-informed selector incorporating these three new layers is
established; and (3) a new recursive iteration algorithm is also proposed to
improve solution optimality. The numerical results on IEEE test systems show
that the proposed method is very effective and promising. |
2,859 | ZeroQuant: Efficient and Affordable Post-Training Quantization for Large-Scale Transformers | How to efficiently serve ever-larger trained natural language models in
practice has become exceptionally challenging even for powerful cloud servers
due to their prohibitive memory/computation requirements. In this work, we
present an efficient and affordable post-training quantization approach to
compress large Transformer-based models, termed as ZeroQuant. ZeroQuant is an
end-to-end quantization and inference pipeline with three main components: (1)
a fine-grained hardware-friendly quantization scheme for both weight and
activations; (2) a novel affordable layer-by-layer knowledge distillation
algorithm (LKD) even without the access to the original training data; (3) a
highly-optimized quantization system backend support to remove the
quantization/dequantization overhead. As such, we are able to show that: (1)
ZeroQuant can reduce the precision for weights and activations to INT8 in a
cost-free way for both BERT and GPT3-style models with minimal accuracy impact,
which leads to up to 5.19x/4.16x speedup on those models compared to FP16
inference; (2) ZeroQuant plus LKD affordably quantize the weights in the
fully-connected module to INT4 along with INT8 weights in the attention module
and INT8 activations, resulting in 3x memory footprint reduction compared to
the FP16 model; (3) ZeroQuant can be directly applied to two of the largest
open-sourced language models, including GPT-J6B and GPT-NeoX20, for which our
INT8 model achieves similar accuracy as the FP16 model but achieves up to 5.2x
better efficiency. |
2,860 | Extreme Compression for Pre-trained Transformers Made Simple and Efficient | Extreme compression, particularly ultra-low bit precision (binary/ternary)
quantization, has been proposed to fit large NLP models on resource-constraint
devices. However, to preserve the accuracy for such aggressive compression
schemes, cutting-edge methods usually introduce complicated compression
pipelines, e.g., multi-stage expensive knowledge distillation with extensive
hyperparameter tuning. Also, they oftentimes focus less on smaller transformer
models that have already been heavily compressed via knowledge distillation and
lack a systematic study to show the effectiveness of their methods. In this
paper, we perform a very comprehensive systematic study to measure the impact
of many key hyperparameters and training strategies from previous works. As a
result, we find out that previous baselines for ultra-low bit precision
quantization are significantly under-trained. Based on our study, we propose a
simple yet effective compression pipeline for extreme compression, named XTC.
XTC demonstrates that (1) we can skip the pre-training knowledge distillation
to obtain a 5-layer BERT while achieving better performance than previous
state-of-the-art methods, e.g., the 6-layer TinyBERT; (2) extreme quantization
plus layer reduction is able to reduce the model size by 50x, resulting in new
state-of-the-art results on GLUE tasks. |
2,861 | Federated Deep Learning Meets Autonomous Vehicle Perception: Design and Verification | Realizing human-like perception is a challenge in open driving scenarios due
to corner cases and visual occlusions. To gather knowledge of rare and occluded
instances, federated learning empowered connected autonomous vehicle (FLCAV)
has been proposed, which leverages vehicular networks to establish federated
deep neural networks (DNNs) from distributed data captured by vehicles and road
sensors. Without the need of data aggregation, FLCAV preserves privacy while
reducing communication and annotation costs compared with conventional
centralized learning. However, it is challenging to determine the network
resources and road sensor poses for multi-stage training with multi-modal
datasets in multi-variant scenarios. This article presents networking and
training frameworks for FLCAV perception. Multi-layer graph resource allocation
and vehicle-road pose contrastive methods are proposed to address the network
management and sensor pose problems, respectively. We also develop CarlaFLCAV,
a software platform that implements the above system and methods. Experimental
results confirm the superiority of the proposed techniques compared with
various benchmarks. |
2,862 | Out-of-Distribution Detection using BiGAN and MDL | We consider the following problem: we have a large dataset of normal data
available. We are now given a new, possibly quite small, set of data, and we
are to decide if these are normal data, or if they are indicating a new
phenomenon. This is a novelty detection or out-of-distribution detection
problem. An example is in medicine, where the normal data is for people with no
known disease, and the new dataset people with symptoms. Other examples could
be in security. We solve this problem by training a bidirectional generative
adversarial network (BiGAN) on the normal data and using a Gaussian graphical
model to model the output. We then use universal source coding, or minimum
description length (MDL) on the output to decide if it is a new distribution,
in an implementation of Kolmogorov and Martin-L\"{o}f randomness. We apply the
methodology to both MNIST data and a real-world electrocardiogram (ECG) dataset
of healthy and patients with Kawasaki disease, and show better performance in
terms of the ROC curve than similar methods. |
2,863 | Coffee Roast Intelligence | As the coffee industry has grown, there would be more demand for roasted
coffee beans, as well as increased rivalry for selling coffee and attracting
customers. As the flavor of each variety of coffee is dependent on the degree
of roasting of the coffee beans, it is vital to maintain a consistent quality
related to the degree of roasting. Each barista has their own method for
determining the degree of roasting. However, extrinsic circumstances such as
light, fatigue, and other factors may alter their judgment. As a result, the
quality of the coffee cannot be controlled. The Coffee Roast Intelligence
application is a machine learning-based study of roasted coffee bean degrees
classification produced as an Android application platform that identifies the
color of coffee beans by photographing or uploading them while roasting. This
application displays the text showing at what level the coffee beans have been
roasted, as well as informs the percent chance of class prediction to the
consumers. Users may also keep track of the result of the predictions related
to the roasting level of coffee beans. |
2,864 | Differentially Private Model Compression | Recent papers have shown that large pre-trained language models (LLMs) such
as BERT, GPT-2 can be fine-tuned on private data to achieve performance
comparable to non-private models for many downstream Natural Language
Processing (NLP) tasks while simultaneously guaranteeing differential privacy.
The inference cost of these models -- which consist of hundreds of millions of
parameters -- however, can be prohibitively large. Hence, often in practice,
LLMs are compressed before they are deployed in specific applications. In this
paper, we initiate the study of differentially private model compression and
propose frameworks for achieving 50% sparsity levels while maintaining nearly
full performance. We demonstrate these ideas on standard GLUE benchmarks using
BERT models, setting benchmarks for future research on this topic. |
2,865 | Dimension Independent Generalization of DP-SGD for Overparameterized Smooth Convex Optimization | This paper considers the generalization performance of differentially private
convex learning. We demonstrate that the convergence analysis of Langevin
algorithms can be used to obtain new generalization bounds with differential
privacy guarantees for DP-SGD. More specifically, by using some recently
obtained dimension-independent convergence results for stochastic Langevin
algorithms with convex objective functions, we obtain $O(n^{-1/4})$ privacy
guarantees for DP-SGD with the optimal excess generalization error of
$\tilde{O}(n^{-1/2})$ for certain classes of overparameterized smooth convex
optimization problems. This improves previous DP-SGD results for such problems
that contain explicit dimension dependencies, so that the resulting
generalization bounds become unsuitable for overparameterized models used in
practical applications. |
2,866 | Drawing out of Distribution with Neuro-Symbolic Generative Models | Learning general-purpose representations from perceptual inputs is a hallmark
of human intelligence. For example, people can write out numbers or characters,
or even draw doodles, by characterizing these tasks as different instantiations
of the same generic underlying process -- compositional arrangements of
different forms of pen strokes. Crucially, learning to do one task, say
writing, implies reasonable competence at another, say drawing, on account of
this shared process. We present Drawing out of Distribution (DooD), a
neuro-symbolic generative model of stroke-based drawing that can learn such
general-purpose representations. In contrast to prior work, DooD operates
directly on images, requires no supervision or expensive test-time inference,
and performs unsupervised amortised inference with a symbolic stroke model that
better enables both interpretability and generalization. We evaluate DooD on
its ability to generalise across both data and tasks. We first perform
zero-shot transfer from one dataset (e.g. MNIST) to another (e.g. Quickdraw),
across five different datasets, and show that DooD clearly outperforms
different baselines. An analysis of the learnt representations further
highlights the benefits of adopting a symbolic stroke model. We then adopt a
subset of the Omniglot challenge tasks, and evaluate its ability to generate
new exemplars (both unconditionally and conditionally), and perform one-shot
classification, showing that DooD matches the state of the art. Taken together,
we demonstrate that DooD does indeed capture general-purpose representations
across both data and task, and takes a further step towards building general
and robust concept-learning systems. |
2,867 | Debiased Machine Learning without Sample-Splitting for Stable Estimators | Estimation and inference on causal parameters is typically reduced to a
generalized method of moments problem, which involves auxiliary functions that
correspond to solutions to a regression or classification problem. Recent line
of work on debiased machine learning shows how one can use generic machine
learning estimators for these auxiliary problems, while maintaining asymptotic
normality and root-$n$ consistency of the target parameter of interest, while
only requiring mean-squared-error guarantees from the auxiliary estimation
algorithms. The literature typically requires that these auxiliary problems are
fitted on a separate sample or in a cross-fitting manner. We show that when
these auxiliary estimation algorithms satisfy natural leave-one-out stability
properties, then sample splitting is not required. This allows for sample
re-use, which can be beneficial in moderately sized sample regimes. For
instance, we show that the stability properties that we propose are satisfied
for ensemble bagged estimators, built via sub-sampling without replacement, a
popular technique in machine learning practice. |
2,868 | A Robust Backpropagation-Free Framework for Images | While current deep learning algorithms have been successful for a wide
variety of artificial intelligence (AI) tasks, including those involving
structured image data, they present deep neurophysiological conceptual issues
due to their reliance on the gradients computed by backpropagation of errors
(backprop) to obtain synaptic weight adjustments; hence are biologically
implausible. We present a more biologically plausible approach, the
error-kernel driven activation alignment (EKDAA) algorithm, to train
convolution neural networks (CNNs) using locally derived error transmission
kernels and error maps. We demonstrate the efficacy of EKDAA by performing the
task of visual-recognition on the Fashion MNIST, CIFAR-10 and SVHN benchmarks
as well as conducting blackbox robustness tests on adversarial examples derived
from these datasets. Furthermore, we also present results for a CNN trained
using a non-differentiable activation function. All recognition results nearly
matches that of backprop and exhibit greater adversarial robustness compared to
backprop. |
2,869 | QAGCN: A Graph Convolutional Network-based Multi-Relation Question Answering System | Answering multi-relation questions over knowledge graphs is a challenging
task as it requires multi-step reasoning over a huge number of possible paths.
Reasoning-based methods with complex reasoning mechanisms, such as
reinforcement learning-based sequential decision making, have been regarded as
the default pathway for this task. However, these mechanisms are difficult to
implement and train, which hampers their reproducibility and transferability to
new domains. In this paper, we propose QAGCN - a simple but effective and novel
model that leverages attentional graph convolutional networks that can perform
multi-step reasoning during the encoding of knowledge graphs. As a consequence,
complex reasoning mechanisms are avoided. In addition, to improve efficiency,
we retrieve answers using highly-efficient embedding computations and, for
better interpretability, we extract interpretable paths for returned answers.
On widely adopted benchmark datasets, the proposed model has been demonstrated
competitive against state-of-the-art methods that rely on complex reasoning
mechanisms. We also conducted extensive experiments to scrutinize the
efficiency and contribution of each component of our model. |
2,870 | Contrastive learning unifies $t$-SNE and UMAP | Neighbor embedding methods $t$-SNE and UMAP are the de facto standard for
visualizing high-dimensional datasets. They appear to use very different loss
functions with different motivations, and the exact relationship between them
has been unclear. Here we show that UMAP is effectively negative sampling
applied to the $t$-SNE loss function. We explain the difference between
negative sampling and noise-contrastive estimation (NCE), which has been used
to optimize $t$-SNE under the name NCVis. We prove that, unlike NCE, negative
sampling learns a scaled data distribution. When applied in the neighbor
embedding setting, it yields more compact embeddings with increased attraction,
explaining differences in appearance between UMAP and $t$-SNE. Further, we
generalize the notion of negative sampling and obtain a spectrum of embeddings,
encompassing visualizations similar to $t$-SNE, NCVis, and UMAP. Finally, we
explore the connection between representation learning in the SimCLR setting
and neighbor embeddings, and show that (i) $t$-SNE can be optimized using the
InfoNCE loss and in a parametric setting; (ii) various contrastive losses with
only few noise samples can yield competitive performance in the SimCLR setup. |
2,871 | Challenges to Solving Combinatorially Hard Long-Horizon Deep RL Tasks | Deep reinforcement learning has shown promise in discrete domains requiring
complex reasoning, including games such as Chess, Go, and Hanabi. However, this
type of reasoning is less often observed in long-horizon, continuous domains
with high-dimensional observations, where instead RL research has predominantly
focused on problems with simple high-level structure (e.g. opening a drawer or
moving a robot as fast as possible). Inspired by combinatorially hard
optimization problems, we propose a set of robotics tasks which admit many
distinct solutions at the high-level, but require reasoning about states and
rewards thousands of steps into the future for the best performance.
Critically, while RL has traditionally suffered on complex, long-horizon tasks
due to sparse rewards, our tasks are carefully designed to be solvable without
specialized exploration. Nevertheless, our investigation finds that standard RL
methods often neglect long-term effects due to discounting, while
general-purpose hierarchical RL approaches struggle unless additional abstract
domain knowledge can be exploited. |
2,872 | Learning Fine Scale Dynamics from Coarse Observations via Inner Recurrence | Recent work has focused on data-driven learning of the evolution of unknown
systems via deep neural networks (DNNs), with the goal of conducting long term
prediction of the dynamics of the unknown system. In many real-world
applications, data from time-dependent systems are often collected on a time
scale that is coarser than desired, due to various restrictions during the data
acquisition process. Consequently, the observed dynamics can be severely
under-sampled and do not reflect the true dynamics of the underlying system.
This paper presents a computational technique to learn the fine-scale dynamics
from such coarsely observed data. The method employs inner recurrence of a DNN
to recover the fine-scale evolution operator of the underlying system. In
addition to mathematical justification, several challenging numerical examples,
including unknown systems of both ordinary and partial differential equations,
are presented to demonstrate the effectiveness of the proposed method. |
2,873 | Do-Operation Guided Causal Representation Learning with Reduced Supervision Strength | Causal representation learning has been proposed to encode relationships
between factors presented in the high dimensional data. However, existing
methods suffer from merely using a large amount of labeled data and ignore the
fact that samples generated by the same causal mechanism follow the same causal
relationships. In this paper, we seek to explore such information by leveraging
do-operation for reducing supervision strength. We propose a framework which
implements do-operation by swapping latent cause and effect factors encoded
from a pair of inputs. Moreover, we also identify the inadequacy of existing
causal representation metrics empirically and theoretically, and introduce new
metrics for better evaluation. Experiments conducted on both synthetic and real
datasets demonstrate the superiorities of our method compared with
state-of-the-art methods. |
2,874 | Robust Topological Inference in the Presence of Outliers | The distance function to a compact set plays a crucial role in the paradigm
of topological data analysis. In particular, the sublevel sets of the distance
function are used in the computation of persistent homology -- a backbone of
the topological data analysis pipeline. Despite its stability to perturbations
in the Hausdorff distance, persistent homology is highly sensitive to outliers.
In this work, we develop a framework of statistical inference for persistent
homology in the presence of outliers. Drawing inspiration from recent
developments in robust statistics, we propose a $\textit{median-of-means}$
variant of the distance function ($\textsf{MoM Dist}$), and establish its
statistical properties. In particular, we show that, even in the presence of
outliers, the sublevel filtrations and weighted filtrations induced by
$\textsf{MoM Dist}$ are both consistent estimators of the true underlying
population counterpart, and their rates of convergence in the bottleneck metric
are controlled by the fraction of outliers in the data. Finally, we demonstrate
the advantages of the proposed methodology through simulations and
applications. |
2,875 | Additive MIL: Intrinsic Interpretability for Pathology | Multiple Instance Learning (MIL) has been widely applied in pathology towards
solving critical problems such as automating cancer diagnosis and grading,
predicting patient prognosis, and therapy response. Deploying these models in a
clinical setting requires careful inspection of these black boxes during
development and deployment to identify failures and maintain physician trust.
In this work, we propose a simple formulation of MIL models, which enables
interpretability while maintaining similar predictive performance. Our Additive
MIL models enable spatial credit assignment such that the contribution of each
region in the image can be exactly computed and visualized. We show that our
spatial credit assignment coincides with regions used by pathologists during
diagnosis and improves upon classical attention heatmaps from attention MIL
models. We show that any existing MIL model can be made additive with a simple
change in function composition. We also show how these models can debug model
failures, identify spurious features, and highlight class-wise regions of
interest, enabling their use in high-stakes environments such as clinical
decision-making. |
2,876 | R2U++: A Multiscale Recurrent Residual U-Net with Dense Skip Connections for Medical Image Segmentation | U-Net is a widely adopted neural network in the domain of medical image
segmentation. Despite its quick embracement by the medical imaging community,
its performance suffers on complicated datasets. The problem can be ascribed to
its simple feature extracting blocks: encoder/decoder, and the semantic gap
between encoder and decoder. Variants of U-Net (such as R2U-Net) have been
proposed to address the problem of simple feature extracting blocks by making
the network deeper, but it does not deal with the semantic gap problem. On the
other hand, another variant UNET++ deals with the semantic gap problem by
introducing dense skip connections but has simple feature extraction blocks. To
overcome these issues, we propose a new U-Net based medical image segmentation
architecture R2U++. In the proposed architecture, the adapted changes from
vanilla U-Net are: (1) the plain convolutional backbone is replaced by a deeper
recurrent residual convolution block. The increased field of view with these
blocks aids in extracting crucial features for segmentation which is proven by
improvement in the overall performance of the network. (2) The semantic gap
between encoder and decoder is reduced by dense skip pathways. These pathways
accumulate features coming from multiple scales and apply concatenation
accordingly. The modified architecture has embedded multi-depth models, and an
ensemble of outputs taken from varying depths improves the performance on
foreground objects appearing at various scales in the images. The performance
of R2U++ is evaluated on four distinct medical imaging modalities: electron
microscopy (EM), X-rays, fundus, and computed tomography (CT). The average gain
achieved in IoU score is 1.5+-0.37% and in dice score is 0.9+-0.33% over
UNET++, whereas, 4.21+-2.72 in IoU and 3.47+-1.89 in dice score over R2U-Net
across different medical imaging segmentation datasets. |
2,877 | Human Activity Recognition on Time Series Accelerometer Sensor Data using LSTM Recurrent Neural Networks | The use of sensors available through smart devices has pervaded everyday life
in several applications including human activity monitoring, healthcare, and
social networks. In this study, we focus on the use of smartwatch accelerometer
sensors to recognize eating activity. More specifically, we collected sensor
data from 10 participants while consuming pizza. Using this information, and
other comparable data available for similar events such as smoking and
medication-taking, and dissimilar activities of jogging, we developed a
LSTM-ANN architecture that has demonstrated 90% success in identifying
individual bites compared to a puff, medication-taking or jogging activities. |
2,878 | Optimal Competitive-Ratio Control | Inspired by competitive policy designs approaches in online learning, new
control paradigms such as competitive-ratio and regret-optimal control have
been recently proposed as alternatives to the classical $\mathcal{H}_2$ and
$\mathcal{H}_\infty$ approaches. These competitive metrics compare the control
cost of the designed controller against the cost of a clairvoyant controller,
which has access to past, present, and future disturbances in terms of ratio
and difference, respectively. While prior work provided the optimal solution
for the regret-optimal control problem, in competitive-ratio control, the
solution is only provided for the sub-optimal problem. In this work, we derive
the optimal solution to the competitive-ratio control problem. We show that the
optimal competitive ratio formula can be computed as the maximal eigenvalue of
a simple matrix, and provide a state-space controller that achieves the optimal
competitive ratio. We conduct an extensive numerical study to verify this
analytical solution, and demonstrate that the optimal competitive-ratio
controller outperforms other controllers on several large scale practical
systems. The key techniques that underpin our explicit solution is a reduction
of the control problem to a Nehari problem, along with a novel factorization of
the clairvoyant controller's cost. We reveal an interesting relation between
the explicit solutions that now exist for both competitive control paradigms by
formulating a regret-optimal control framework with weight functions that can
also be utilized for practical purposes. |
2,879 | A Learning-Based Method for Automatic Operator Selection in the Fanoos XAI System | We describe an extension of the Fanoos XAI system [Bayani et al 2022] which
enables the system to learn the appropriate action to take in order to satisfy
a user's request for description to be made more or less abstract.
Specifically, descriptions of systems under analysis are stored in states, and
in order to make a description more or less abstract, Fanoos selects an
operator from a large library to apply to the state and generate a new
description. Prior work on Fanoos predominately used hand-written methods for
operator-selection; this current work allows Fanoos to leverage experience to
learn the best operator to apply in a particular situation, balancing
exploration and exploitation, leveraging expert insights when available, and
utilizing similarity between the current state and past states. Additionally,
in order to bootstrap the learning process (i.e., like in curriculum learning),
we describe a simulated user which we implemented; this simulation allows
Fanoos to gain general insights that enable reasonable courses of action,
insights which later can be refined by experience with real users, as opposed
to interacting with humans completely from scratch. Code implementing the
methods described in the paper can be found at
https://github/DBay-ani/Operator_Selection_Learning_Extensions_For_Fanoos. |
2,880 | Revisiting the "Video" in Video-Language Understanding | What makes a video task uniquely suited for videos, beyond what can be
understood from a single image? Building on recent progress in self-supervised
image-language models, we revisit this question in the context of video and
language tasks. We propose the atemporal probe (ATP), a new model for
video-language analysis which provides a stronger bound on the baseline
accuracy of multimodal models constrained by image-level understanding. By
applying this model to standard discriminative video and language tasks, such
as video question answering and text-to-video retrieval, we characterize the
limitations and potential of current video-language benchmarks. We find that
understanding of event temporality is often not necessary to achieve strong or
state-of-the-art performance, even compared with recent large-scale
video-language models and in contexts intended to benchmark deeper video-level
understanding. We also demonstrate how ATP can improve both video-language
dataset and model design. We describe a technique for leveraging ATP to better
disentangle dataset subsets with a higher concentration of temporally
challenging data, improving benchmarking efficacy for causal and temporal
understanding. Further, we show that effectively integrating ATP into full
video-level temporal models can improve efficiency and state-of-the-art
accuracy. |
2,881 | A Theoretical Analysis on Feature Learning in Neural Networks: Emergence from Inputs and Advantage over Fixed Features | An important characteristic of neural networks is their ability to learn
representations of the input data with effective features for prediction, which
is believed to be a key factor to their superior empirical performance. To
better understand the source and benefit of feature learning in neural
networks, we consider learning problems motivated by practical data, where the
labels are determined by a set of class relevant patterns and the inputs are
generated from these along with some background patterns. We prove that neural
networks trained by gradient descent can succeed on these problems. The success
relies on the emergence and improvement of effective features, which are
learned among exponentially many candidates efficiently by exploiting the data
(in particular, the structure of the input distribution). In contrast, no
linear models on data-independent features of polynomial sizes can learn to as
good errors. Furthermore, if the specific input structure is removed, then no
polynomial algorithm in the Statistical Query model can learn even weakly.
These results provide theoretical evidence showing that feature learning in
neural networks depends strongly on the input structure and leads to the
superior performance. Our preliminary experimental results on synthetic and
real data also provide positive support. |
2,882 | Towards Evading the Limits of Randomized Smoothing: A Theoretical Analysis | Randomized smoothing is the dominant standard for provable defenses against
adversarial examples. Nevertheless, this method has recently been proven to
suffer from important information theoretic limitations. In this paper, we
argue that these limitations are not intrinsic, but merely a byproduct of
current certification methods. We first show that these certificates use too
little information about the classifier, and are in particular blind to the
local curvature of the decision boundary. This leads to severely sub-optimal
robustness guarantees as the dimension of the problem increases. We then show
that it is theoretically possible to bypass this issue by collecting more
information about the classifier. More precisely, we show that it is possible
to approximate the optimal certificate with arbitrary precision, by probing the
decision boundary with several noise distributions. Since this process is
executed at certification time rather than at test time, it entails no loss in
natural accuracy while enhancing the quality of the certificates. This result
fosters further research on classifier-specific certification and demonstrates
that randomized smoothing is still worth investigating. Although
classifier-specific certification may induce more computational cost, we also
provide some theoretical insight on how to mitigate it. |
2,883 | Compositional Visual Generation with Composable Diffusion Models | Large text-guided diffusion models, such as DALLE-2, are able to generate
stunning photorealistic images given natural language descriptions. While such
models are highly flexible, they struggle to understand the composition of
certain concepts, such as confusing the attributes of different objects or
relations between objects. In this paper, we propose an alternative structured
approach for compositional generation using diffusion models. An image is
generated by composing a set of diffusion models, with each of them modeling a
certain component of the image. To do this, we interpret diffusion models as
energy-based models in which the data distributions defined by the energy
functions may be explicitly combined. The proposed method can generate scenes
at test time that are substantially more complex than those seen in training,
composing sentence descriptions, object relations, human facial attributes, and
even generalizing to new combinations that are rarely seen in the real world.
We further illustrate how our approach may be used to compose pre-trained
text-guided diffusion models and generate photorealistic images containing all
the details described in the input descriptions, including the binding of
certain object attributes that have been shown difficult for DALLE-2. These
results point to the effectiveness of the proposed method in promoting
structured generalization for visual generation. Project page:
https://energy-based-model.github.io/Compositional-Visual-Generation-with-Composable-Diffusion-Models/ |
2,884 | KCRL: Krasovskii-Constrained Reinforcement Learning with Guaranteed Stability in Nonlinear Dynamical Systems | Learning a dynamical system requires stabilizing the unknown dynamics to
avoid state blow-ups. However, current reinforcement learning (RL) methods lack
stabilization guarantees, which limits their applicability for the control of
safety-critical systems. We propose a model-based RL framework with formal
stability guarantees, Krasovskii Constrained RL (KCRL), that adopts
Krasovskii's family of Lyapunov functions as a stability constraint. The
proposed method learns the system dynamics up to a confidence interval using
feature representation, e.g. Random Fourier Features. It then solves a
constrained policy optimization problem with a stability constraint based on
Krasovskii's method using a primal-dual approach to recover a stabilizing
policy. We show that KCRL is guaranteed to learn a stabilizing policy in a
finite number of interactions with the underlying unknown system. We also
derive the sample complexity upper bound for stabilization of unknown nonlinear
dynamical systems via the KCRL framework. |
2,885 | Scalar is Not Enough: Vectorization-based Unbiased Learning to Rank | Unbiased learning to rank (ULTR) aims to train an unbiased ranking model from
biased user click logs. Most of the current ULTR methods are based on the
examination hypothesis (EH), which assumes that the click probability can be
factorized into two scalar functions, one related to ranking features and the
other related to bias factors. Unfortunately, the interactions among features,
bias factors and clicks are complicated in practice, and usually cannot be
factorized in this independent way. Fitting click data with EH could lead to
model misspecification and bring the approximation error.
In this paper, we propose a vector-based EH and formulate the click
probability as a dot product of two vector functions. This solution is complete
due to its universality in fitting arbitrary click functions. Based on it, we
propose a novel model named Vectorization to adaptively learn the relevance
embeddings and sort documents by projecting embeddings onto a base vector.
Extensive experiments show that our method significantly outperforms the
state-of-the-art ULTR methods on complex real clicks as well as simple
simulated clicks. |
2,886 | Deep Learning Prediction of Severe Health Risks for Pediatric COVID-19 Patients with a Large Feature Set in 2021 BARDA Data Challenge | Most children infected with COVID-19 have no or mild symptoms and can recover
automatically by themselves, but some pediatric COVID-19 patients need to be
hospitalized or even to receive intensive medical care (e.g., invasive
mechanical ventilation or cardiovascular support) to recover from the
illnesses. Therefore, it is critical to predict the severe health risk that
COVID-19 infection poses to children to provide precise and timely medical care
for vulnerable pediatric COVID-19 patients. However, predicting the severe
health risk for COVID-19 patients including children remains a significant
challenge because many underlying medical factors affecting the risk are still
largely unknown. In this work, instead of searching for a small number of most
useful features to make prediction, we design a novel large-scale bag-of-words
like method to represent various medical conditions and measurements of
COVID-19 patients. After some simple feature filtering based on logistical
regression, the large set of features is used with a deep learning method to
predict both the hospitalization risk for COVID-19 infected children and the
severe complication risk for the hospitalized pediatric COVID-19 patients. The
method was trained and tested the datasets of the Biomedical Advanced Research
and Development Authority (BARDA) Pediatric COVID-19 Data Challenge held from
Sept. 15 to Dec. 17, 2021. The results show that the approach can rather
accurately predict the risk of hospitalization and severe complication for
pediatric COVID-19 patients and deep learning is more accurate than other
machine learning methods. |
2,887 | Measuring Gender Bias in Word Embeddings of Gendered Languages Requires Disentangling Grammatical Gender Signals | Does the grammatical gender of a language interfere when measuring the
semantic gender information captured by its word embeddings? A number of
anomalous gender bias measurements in the embeddings of gendered languages
suggest this possibility. We demonstrate that word embeddings learn the
association between a noun and its grammatical gender in grammatically gendered
languages, which can skew social gender bias measurements. Consequently, word
embedding post-processing methods are introduced to quantify, disentangle, and
evaluate grammatical gender signals. The evaluation is performed on five
gendered languages from the Germanic, Romance, and Slavic branches of the
Indo-European language family. Our method reduces the strength of grammatical
gender signals, which is measured in terms of effect size (Cohen's d), by a
significant average of d = 1.3 for French, German, and Italian, and d = 0.56
for Polish and Spanish. Once grammatical gender is disentangled, the
association between over 90% of 10,000 inanimate nouns and their assigned
grammatical gender weakens, and cross-lingual bias results from the Word
Embedding Association Test (WEAT) become more congruent with country-level
implicit bias measurements. The results further suggest that disentangling
grammatical gender signals from word embeddings may lead to improvement in
semantic machine learning tasks. |
2,888 | Dynamic Kernel Selection for Improved Generalization and Memory Efficiency in Meta-learning | Gradient based meta-learning methods are prone to overfit on the
meta-training set, and this behaviour is more prominent with large and complex
networks. Moreover, large networks restrict the application of meta-learning
models on low-power edge devices. While choosing smaller networks avoid these
issues to a certain extent, it affects the overall generalization leading to
reduced performance. Clearly, there is an approximately optimal choice of
network architecture that is best suited for every meta-learning problem,
however, identifying it beforehand is not straightforward. In this paper, we
present MetaDOCK, a task-specific dynamic kernel selection strategy for
designing compressed CNN models that generalize well on unseen tasks in
meta-learning. Our method is based on the hypothesis that for a given set of
similar tasks, not all kernels of the network are needed by each individual
task. Rather, each task uses only a fraction of the kernels, and the selection
of the kernels per task can be learnt dynamically as a part of the inner update
steps. MetaDOCK compresses the meta-model as well as the task-specific inner
models, thus providing significant reduction in model size for each task, and
through constraining the number of active kernels for every task, it implicitly
mitigates the issue of meta-overfitting. We show that for the same inference
budget, pruned versions of large CNN models obtained using our approach
consistently outperform the conventional choices of CNN models. MetaDOCK
couples well with popular meta-learning approaches such as iMAML. The efficacy
of our method is validated on CIFAR-fs and mini-ImageNet datasets, and we have
observed that our approach can provide improvements in model accuracy of up to
2% on standard meta-learning benchmark, while reducing the model size by more
than 75%. |
2,889 | Algorithm for Constrained Markov Decision Process with Linear Convergence | The problem of constrained Markov decision process is considered. An agent
aims to maximize the expected accumulated discounted reward subject to multiple
constraints on its costs (the number of constraints is relatively small). A new
dual approach is proposed with the integration of two ingredients: entropy
regularized policy optimizer and Vaidya's dual optimizer, both of which are
critical to achieve faster convergence. The finite-time error bound of the
proposed approach is provided. Despite the challenge of the nonconcave
objective subject to nonconcave constraints, the proposed approach is shown to
converge (with linear rate) to the global optimum. The complexity expressed in
terms of the optimality gap and the constraint violation significantly improves
upon the existing primal-dual approaches. |
2,890 | BaCaDI: Bayesian Causal Discovery with Unknown Interventions | Learning causal structures from observation and experimentation is a central
task in many domains. For example, in biology, recent advances allow us to
obtain single-cell expression data under multiple interventions such as drugs
or gene knockouts. However, a key challenge is that often the targets of the
interventions are uncertain or unknown. Thus, standard causal discovery methods
can no longer be used. To fill this gap, we propose a Bayesian framework
(BaCaDI) for discovering the causal structure that underlies data generated
under various unknown experimental/interventional conditions. BaCaDI is fully
differentiable and operates in the continuous space of latent probabilistic
representations of both causal structures and interventions. This enables us to
approximate complex posteriors via gradient-based variational inference and to
reason about the epistemic uncertainty in the predicted structure. In
experiments on synthetic causal discovery tasks and simulated gene-expression
data, BaCaDI outperforms related methods in identifying causal structures and
intervention targets. Finally, we demonstrate that, thanks to its rigorous
Bayesian approach, our method provides well-calibrated uncertainty estimates. |
2,891 | Joint Energy Dispatch and Unit Commitment in Microgrids Based on Deep Reinforcement Learning | Nowadays, the application of microgrids (MG) with renewable energy is
becoming more and more extensive, which creates a strong need for dynamic
energy management. In this paper, deep reinforcement learning (DRL) is applied
to learn an optimal policy for making joint energy dispatch (ED) and unit
commitment (UC) decisions in an isolated MG, with the aim for reducing the
total power generation cost on the premise of ensuring the supply-demand
balance. In order to overcome the challenge of discrete-continuous hybrid
action space due to joint ED and UC, we propose a DRL algorithm, i.e., the
hybrid action finite-horizon DDPG (HAFH-DDPG), that seamlessly integrates two
classical DRL algorithms, i.e., deep Q-network (DQN) and deep deterministic
policy gradient (DDPG), based on a finite-horizon dynamic programming (DP)
framework. Moreover, a diesel generator (DG) selection strategy is presented to
support a simplified action space for reducing the computation complexity of
this algorithm. Finally, the effectiveness of our proposed algorithm is
verified through comparison with several baseline algorithms by experiments
with real-world data set. |
2,892 | Uncertainty Estimation in Machine Learning | Most machine learning techniques are based upon statistical learning theory,
often simplified for the sake of computing speed. This paper is focused on the
uncertainty aspect of mathematical modeling in machine learning. Regression
analysis is chosen to further investigate the evaluation aspect of uncertainty
in model coefficients and, more importantly, in the output feature value
predictions. A survey demonstrates major stages in the conventional least
squares approach to the creation of the regression model, along with its
uncertainty estimation. On the other hand, it is shown that in machine learning
the model complexity and severe nonlinearity become serious obstacles to
uncertainty evaluation. Furthermore, the process of machine model training
demands high computing power, not available at the level of personal computers.
This is why so-called pre-trained models are widely used in such areas of
machine learning as natural language processing. The latest example of a
pre-trained model is the Generative Pre-trained Transformer 3 with hundreds of
billions of parameters and a half-terabyte training dataset. Similarly,
mathematical models built from real data are growing in complexity which is
accompanied by the growing amount of training data. However, when machine
models and their predictions are used in decision-making, one needs to estimate
uncertainty and evaluate accompanying risks. This problem could be resolved
with non-parametric techniques at the expense of greater demand for computing
power, which can be offered by modern supercomputers available, including those
utilizing graphical and tensor processing units along with the conventional
central processors. |
2,893 | Neural Differential Equations for Learning to Program Neural Nets Through Continuous Learning Rules | Neural ordinary differential equations (ODEs) have attracted much attention
as continuous-time counterparts of deep residual neural networks (NNs), and
numerous extensions for recurrent NNs have been proposed. Since the 1980s, ODEs
have also been used to derive theoretical results for NN learning rules, e.g.,
the famous connection between Oja's rule and principal component analysis. Such
rules are typically expressed as additive iterative update processes which have
straightforward ODE counterparts. Here we introduce a novel combination of
learning rules and Neural ODEs to build continuous-time sequence processing
nets that learn to manipulate short-term memory in rapidly changing synaptic
connections of other nets. This yields continuous-time counterparts of Fast
Weight Programmers and linear Transformers. Our novel models outperform the
best existing Neural Controlled Differential Equation based models on various
time series classification tasks, while also addressing their scalability
limitations. Our code is public. |
2,894 | Multi-user Co-inference with Batch Processing Capable Edge Server | Graphics processing units (GPUs) can improve deep neural network inference
throughput via batch processing, where multiple tasks are concurrently
processed. We focus on novel scenarios that the energy-constrained mobile
devices offload inference tasks to an edge server with GPU. The inference task
is partitioned into sub-tasks for a finer granularity of offloading and
scheduling, and the user energy consumption minimization problem under
inference latency constraints is investigated. To deal with the coupled
offloading and scheduling introduced by concurrent batch processing, we first
consider an offline problem with a constant edge inference latency and the same
latency constraint. It is proven that optimizing the offloading policy of each
user independently and aggregating all the same sub-tasks in one batch is
optimal, and thus the independent partitioning and same sub-task aggregating
(IP-SSA) algorithm is inspired. Further, the optimal grouping (OG) algorithm is
proposed to optimally group tasks when the latency constraints are different.
Finally, when future task arrivals cannot be precisely predicted, a deep
deterministic policy gradient (DDPG) agent is trained to call OG. Experiments
show that IP-SSA reduces up to 94.9\% user energy consumption in the offline
setting, while DDPG-OG outperforms DDPG-IP-SSA by up to 8.92\% in the online
setting. |
2,895 | PROMISSING: Pruning Missing Values in Neural Networks | While data are the primary fuel for machine learning models, they often
suffer from missing values, especially when collected in real-world scenarios.
However, many off-the-shelf machine learning models, including artificial
neural network models, are unable to handle these missing values directly.
Therefore, extra data preprocessing and curation steps, such as data
imputation, are inevitable before learning and prediction processes. In this
study, we propose a simple and intuitive yet effective method for pruning
missing values (PROMISSING) during learning and inference steps in neural
networks. In this method, there is no need to remove or impute the missing
values; instead, the missing values are treated as a new source of information
(representing what we do not know). Our experiments on simulated data, several
classification and regression benchmarks, and a multi-modal clinical dataset
show that PROMISSING results in similar prediction performance compared to
various imputation techniques. In addition, our experiments show models trained
using PROMISSING techniques are becoming less decisive in their predictions
when facing incomplete samples with many unknowns. This finding hopefully
advances machine learning models from being pure predicting machines to more
realistic thinkers that can also say "I do not know" when facing incomplete
sources of information. |
2,896 | Reinforcement Learning with Neural Radiance Fields | It is a long-standing problem to find effective representations for training
reinforcement learning (RL) agents. This paper demonstrates that learning state
representations with supervision from Neural Radiance Fields (NeRFs) can
improve the performance of RL compared to other learned representations or even
low-dimensional, hand-engineered state information. Specifically, we propose to
train an encoder that maps multiple image observations to a latent space
describing the objects in the scene. The decoder built from a
latent-conditioned NeRF serves as the supervision signal to learn the latent
space. An RL algorithm then operates on the learned latent space as its state
representation. We call this NeRF-RL. Our experiments indicate that NeRF as
supervision leads to a latent space better suited for the downstream RL tasks
involving robotic object manipulations like hanging mugs on hooks, pushing
objects, or opening doors. Video: https://dannydriess.github.io/nerf-rl |
2,897 | Pruning for Interpretable, Feature-Preserving Circuits in CNNs | Deep convolutional neural networks are a powerful model class for a range of
computer vision problems, but it is difficult to interpret the image filtering
process they implement, given their sheer size. In this work, we introduce a
method for extracting 'feature-preserving circuits' from deep CNNs, leveraging
methods from saliency-based neural network pruning. These circuits are modular
sub-functions, embedded within the network, containing only a subset of
convolutional kernels relevant to a target feature. We compare the efficacy of
3 saliency-criteria for extracting these sparse circuits. Further, we show how
'sub-feature' circuits can be extracted, that preserve a feature's responses to
particular images, dividing the feature into even sparser filtering processes.
We also develop a tool for visualizing 'circuit diagrams', which render the
entire image filtering process implemented by circuits in a parsable format. |
2,898 | Beyond Tabula Rasa: Reincarnating Reinforcement Learning | Learning tabula rasa, that is without any prior knowledge, is the prevalent
workflow in reinforcement learning (RL) research. However, RL systems, when
applied to large-scale settings, rarely operate tabula rasa. Such large-scale
systems undergo multiple design or algorithmic changes during their development
cycle and use ad hoc approaches for incorporating these changes without
re-training from scratch, which would have been prohibitively expensive.
Additionally, the inefficiency of deep RL typically excludes researchers
without access to industrial-scale resources from tackling
computationally-demanding problems. To address these issues, we present
reincarnating RL as an alternative workflow, where prior computational work
(e.g., learned policies) is reused or transferred between design iterations of
an RL agent, or from one RL agent to another. As a step towards enabling
reincarnating RL from any agent to any other agent, we focus on the specific
setting of efficiently transferring an existing sub-optimal policy to a
standalone value-based RL agent. We find that existing approaches fail in this
setting and propose a simple algorithm to address their limitations. Equipped
with this algorithm, we demonstrate reincarnating RL's gains over tabula rasa
RL on Atari 2600 games, a challenging locomotion task, and the real-world
problem of navigating stratospheric balloons. Overall, this work argues for an
alternative approach to RL research, which we believe could significantly
improve real-world RL adoption and help democratize it further. |
2,899 | Effects of Auxiliary Knowledge on Continual Learning | In Continual Learning (CL), a neural network is trained on a stream of data
whose distribution changes over time. In this context, the main problem is how
to learn new information without forgetting old knowledge (i.e., Catastrophic
Forgetting). Most existing CL approaches focus on finding solutions to preserve
acquired knowledge, so working on the past of the model. However, we argue that
as the model has to continually learn new tasks, it is also important to put
focus on the present knowledge that could improve following tasks learning. In
this paper we propose a new, simple, CL algorithm that focuses on solving the
current task in a way that might facilitate the learning of the next ones. More
specifically, our approach combines the main data stream with a secondary,
diverse and uncorrelated stream, from which the network can draw auxiliary
knowledge. This helps the model from different perspectives, since auxiliary
data may contain useful features for the current and the next tasks and
incoming task classes can be mapped onto auxiliary classes. Furthermore, the
addition of data to the current task is implicitly making the classifier more
robust as we are forcing the extraction of more discriminative features. Our
method can outperform existing state-of-the-art models on the most common CL
Image Classification benchmarks. |
Subsets and Splits
No saved queries yet
Save your SQL queries to embed, download, and access them later. Queries will appear here once saved.