paper_id
stringlengths
9
24
title
stringlengths
10
192
paper_url
stringlengths
44
59
authors
listlengths
1
23
type
stringclasses
0 values
primary_area
stringclasses
0 values
abstract
large_stringlengths
82
4.49k
keywords
listlengths
0
0
TL;DR
large_stringclasses
0 values
submission_number
int64
1
1.18k
arxiv_id
stringlengths
10
10
arxiv_id_source
stringclasses
2 values
v139/hu21d
Generalizable Episodic Memory for Deep Reinforcement Learning
https://proceedings.mlr.press/v139/hu21d.html
[ "Hao Hu", "Jianing Ye", "Guangxiang Zhu", "Zhizhou Ren", "Chongjie Zhang" ]
null
null
Episodic memory-based methods can rapidly latch onto past successful strategies by a non-parametric memory and improve sample efficiency of traditional reinforcement learning. However, little effort is put into the continuous domain, where a state is never visited twice, and previous episodic methods fail to efficientl...
[]
null
401
2103.06469
title_snapshot
v139/hua21a
A Scalable Deterministic Global Optimization Algorithm for Clustering Problems
https://proceedings.mlr.press/v139/hua21a.html
[ "Kaixun Hua", "Mingfei Shi", "Yankai Cao" ]
null
null
The minimum sum-of-squares clustering (MSSC) task, which can be treated as a Mixed Integer Second Order Cone Programming (MISOCP) problem, is rarely investigated in the literature through deterministic optimization to find its global optimal value. In this paper, we modelled the MSSC task as a two-stage optimization pr...
[]
null
402
null
null
v139/huang21a
On Recovering from Modeling Errors Using Testing Bayesian Networks
https://proceedings.mlr.press/v139/huang21a.html
[ "Haiying Huang", "Adnan Darwiche" ]
null
null
We consider the problem of supervised learning with Bayesian Networks when the used dependency structure is incomplete due to missing edges or missing variable states. These modeling errors induce independence constraints on the learned model that may not hold in the true, data-generating distribution. We provide a uni...
[]
null
403
null
null
v139/huang21b
A Novel Sequential Coreset Method for Gradient Descent Algorithms
https://proceedings.mlr.press/v139/huang21b.html
[ "Jiawei Huang", "Ruomin Huang", "Wenjie Liu", "Nikolaos Freris", "Hu Ding" ]
null
null
A wide range of optimization problems arising in machine learning can be solved by gradient descent algorithms, and a central question in this area is how to efficiently compress a large-scale dataset so as to reduce the computational complexity. Coreset is a popular data compression technique that has been extensively...
[]
null
404
2112.02504
title_snapshot
v139/huang21c
FL-NTK: A Neural Tangent Kernel-based Framework for Federated Learning Analysis
https://proceedings.mlr.press/v139/huang21c.html
[ "Baihe Huang", "Xiaoxiao Li", "Zhao Song", "Xin Yang" ]
null
null
Federated Learning (FL) is an emerging learning scheme that allows different distributed clients to train deep neural networks together without data sharing. Neural networks have become popular due to their unprecedented success. To the best of our knowledge, the theoretical guarantees of FL concerning neural networks ...
[]
null
405
2105.05001
title_judge
v139/huang21d
STRODE: Stochastic Boundary Ordinary Differential Equation
https://proceedings.mlr.press/v139/huang21d.html
[ "Hengguan Huang", "Hongfu Liu", "Hao Wang", "Chang Xiao", "Ye Wang" ]
null
null
Perception of time from sequentially acquired sensory inputs is rooted in everyday behaviors of individual organisms. Yet, most algorithms for time-series modeling fail to learn dynamics of random event timings directly from visual or audio inputs, requiring timing annotations during training that are usually unavailab...
[]
null
406
2107.08273
title_snapshot
v139/huang21e
A Riemannian Block Coordinate Descent Method for Computing the Projection Robust Wasserstein Distance
https://proceedings.mlr.press/v139/huang21e.html
[ "Minhui Huang", "Shiqian Ma", "Lifeng Lai" ]
null
null
The Wasserstein distance has become increasingly important in machine learning and deep learning. Despite its popularity, the Wasserstein distance is hard to approximate because of the curse of dimensionality. A recently proposed approach to alleviate the curse of dimensionality is to project the sampled data from the ...
[]
null
407
2012.05199
title_snapshot
v139/huang21f
Projection Robust Wasserstein Barycenters
https://proceedings.mlr.press/v139/huang21f.html
[ "Minhui Huang", "Shiqian Ma", "Lifeng Lai" ]
null
null
Collecting and aggregating information from several probability measures or histograms is a fundamental task in machine learning. One of the popular solution methods for this task is to compute the barycenter of the probability measures under the Wasserstein metric. However, approximating the Wasserstein barycenter is ...
[]
null
408
2102.03390
title_snapshot
v139/hubara21a
Accurate Post Training Quantization With Small Calibration Sets
https://proceedings.mlr.press/v139/hubara21a.html
[ "Itay Hubara", "Yury Nahshan", "Yair Hanani", "Ron Banner", "Daniel Soudry" ]
null
null
Lately, post-training quantization methods have gained considerable attention, as they are simple to use, and require only a small unlabeled calibration set. This small dataset cannot be used to fine-tune the model without significant over-fitting. Instead, these methods only use the calibration set to set the activati...
[]
null
409
null
null
v139/hubert21a
Learning and Planning in Complex Action Spaces
https://proceedings.mlr.press/v139/hubert21a.html
[ "Thomas Hubert", "Julian Schrittwieser", "Ioannis Antonoglou", "Mohammadamin Barekatain", "Simon Schmitt", "David Silver" ]
null
null
Many important real-world problems have action spaces that are high-dimensional, continuous or both, making full enumeration of all possible actions infeasible. Instead, only small subsets of actions can be sampled for the purpose of policy evaluation and improvement. In this paper, we propose a general framework to re...
[]
null
410
2104.06303
title_snapshot
v139/hudson21a
Generative Adversarial Transformers
https://proceedings.mlr.press/v139/hudson21a.html
[ "Drew A Hudson", "Larry Zitnick" ]
null
null
We introduce the GANsformer, a novel and efficient type of transformer, and explore it for the task of visual generative modeling. The network employs a bipartite structure that enables long-range interactions across the image, while maintaining computation of linear efficiency, that can readily scale to high-resolutio...
[]
null
411
2103.01209
title_snapshot
v139/hussain21a
Neural Pharmacodynamic State Space Modeling
https://proceedings.mlr.press/v139/hussain21a.html
[ "Zeshan M Hussain", "Rahul G. Krishnan", "David Sontag" ]
null
null
Modeling the time-series of high-dimensional, longitudinal data is important for predicting patient disease progression. However, existing neural network based approaches that learn representations of patient state, while very flexible, are susceptible to overfitting. We propose a deep generative model that makes use o...
[]
null
412
2102.11218
title_snapshot
v139/hussenot21a
Hyperparameter Selection for Imitation Learning
https://proceedings.mlr.press/v139/hussenot21a.html
[ "Léonard Hussenot", "Marcin Andrychowicz", "Damien Vincent", "Robert Dadashi", "Anton Raichuk", "Sabela Ramos", "Nikola Momchev", "Sertan Girgin", "Raphael Marinier", "Lukasz Stafiniak", "Manu Orsini", "Olivier Bachem", "Matthieu Geist", "Olivier Pietquin" ]
null
null
We address the issue of tuning hyperparameters (HPs) for imitation learning algorithms in the context of continuous-control, when the underlying reward function of the demonstrating expert cannot be observed at any time. The vast literature in imitation learning mostly considers this reward function to be available for...
[]
null
413
2105.12034
title_snapshot
v139/huster21a
Pareto GAN: Extending the Representational Power of GANs to Heavy-Tailed Distributions
https://proceedings.mlr.press/v139/huster21a.html
[ "Todd Huster", "Jeremy Cohen", "Zinan Lin", "Kevin Chan", "Charles Kamhoua", "Nandi O. Leslie", "Cho-Yu Jason Chiang", "Vyas Sekar" ]
null
null
Generative adversarial networks (GANs) are often billed as "universal distribution learners", but precisely what distributions they can represent and learn is still an open question. Heavy-tailed distributions are prevalent in many different domains such as financial risk-assessment, physics, and epidemiology. We obser...
[]
null
414
2101.09113
title_snapshot
v139/hutchinson21a
LieTransformer: Equivariant Self-Attention for Lie Groups
https://proceedings.mlr.press/v139/hutchinson21a.html
[ "Michael J Hutchinson", "Charline Le Lan", "Sheheryar Zaidi", "Emilien Dupont", "Yee Whye Teh", "Hyunjik Kim" ]
null
null
Group equivariant neural networks are used as building blocks of group invariant neural networks, which have been shown to improve generalisation performance and data efficiency through principled parameter sharing. Such works have mostly focused on group equivariant convolutions, building on the result that group equi...
[]
null
415
2012.10885
title_snapshot
v139/ibrahim21a
Crowdsourcing via Annotator Co-occurrence Imputation and Provable Symmetric Nonnegative Matrix Factorization
https://proceedings.mlr.press/v139/ibrahim21a.html
[ "Shahana Ibrahim", "Xiao Fu" ]
null
null
Unsupervised learning of the Dawid-Skene (D&S) model from noisy, incomplete and crowdsourced annotations has been a long-standing challenge, and is a critical step towards reliably labeling massive data. A recent work takes a coupled nonnegative matrix factorization (CNMF) perspective, and shows appealing features: It ...
[]
null
416
2106.07193
title_snapshot
v139/ilse21a
Selecting Data Augmentation for Simulating Interventions
https://proceedings.mlr.press/v139/ilse21a.html
[ "Maximilian Ilse", "Jakub M Tomczak", "Patrick Forré" ]
null
null
Machine learning models trained with purely observational data and the principle of empirical risk minimization (Vapnik 1992) can fail to generalize to unseen domains. In this paper, we focus on the case where the problem arises through spurious correlation between the observed domains and the actual task labels. We fi...
[]
null
417
2005.01856
title_snapshot
v139/immer21a
Scalable Marginal Likelihood Estimation for Model Selection in Deep Learning
https://proceedings.mlr.press/v139/immer21a.html
[ "Alexander Immer", "Matthias Bauer", "Vincent Fortuin", "Gunnar Rätsch", "Khan Mohammad Emtiyaz" ]
null
null
Marginal-likelihood based model-selection, even though promising, is rarely used in deep learning due to estimation difficulties. Instead, most approaches rely on validation data, which may not be readily available. In this work, we present a scalable marginal-likelihood estimation method to select both hyperparameters...
[]
null
418
2104.04975
title_snapshot
v139/inatsu21a
Active Learning for Distributionally Robust Level-Set Estimation
https://proceedings.mlr.press/v139/inatsu21a.html
[ "Yu Inatsu", "Shogo Iwazaki", "Ichiro Takeuchi" ]
null
null
Many cases exist in which a black-box function $f$ with high evaluation cost depends on two types of variables $\bm x$ and $\bm w$, where $\bm x$ is a controllable \emph{design} variable and $\bm w$ are uncontrollable \emph{environmental} variables that have random variation following a certain distribution $P$. In suc...
[]
null
419
2102.04000
title_snapshot
v139/indelman21a
Learning Randomly Perturbed Structured Predictors for Direct Loss Minimization
https://proceedings.mlr.press/v139/indelman21a.html
[ "Hedda Cohen Indelman", "Tamir Hazan" ]
null
null
Direct loss minimization is a popular approach for learning predictors over structured label spaces. This approach is computationally appealing as it replaces integration with optimization and allows to propagate gradients in a deep net using loss-perturbed prediction. Recently, this technique was extended to generativ...
[]
null
420
2007.05724
title_snapshot
v139/iqbal21a
Randomized Entity-wise Factorization for Multi-Agent Reinforcement Learning
https://proceedings.mlr.press/v139/iqbal21a.html
[ "Shariq Iqbal", "Christian A Schroeder De Witt", "Bei Peng", "Wendelin Boehmer", "Shimon Whiteson", "Fei Sha" ]
null
null
Multi-agent settings in the real world often involve tasks with varying types and quantities of agents and non-agent entities; however, common patterns of behavior often emerge among these agents/entities. Our method aims to leverage these commonalities by asking the question: “What is the expected utility of each agen...
[]
null
421
2006.04222
title_snapshot
v139/ishfaq21a
Randomized Exploration in Reinforcement Learning with General Value Function Approximation
https://proceedings.mlr.press/v139/ishfaq21a.html
[ "Haque Ishfaq", "Qiwen Cui", "Viet Nguyen", "Alex Ayoub", "Zhuoran Yang", "Zhaoran Wang", "Doina Precup", "Lin Yang" ]
null
null
We propose a model-free reinforcement learning algorithm inspired by the popular randomized least squares value iteration (RLSVI) algorithm as well as the optimism principle. Unlike existing upper-confidence-bound (UCB) based approaches, which are often computationally intractable, our algorithm drives exploration by s...
[]
null
422
2106.07841
title_judge
v139/islamov21a
Distributed Second Order Methods with Fast Rates and Compressed Communication
https://proceedings.mlr.press/v139/islamov21a.html
[ "Rustem Islamov", "Xun Qian", "Peter Richtarik" ]
null
null
We develop several new communication-efficient second-order methods for distributed optimization. Our first method, NEWTON-STAR, is a variant of Newton’s method from which it inherits its fast local quadratic rate. However, unlike Newton’s method, NEWTON-STAR enjoys the same per iteration communication cost as gradient...
[]
null
423
2102.07158
title_snapshot
v139/izmailov21a
What Are Bayesian Neural Network Posteriors Really Like?
https://proceedings.mlr.press/v139/izmailov21a.html
[ "Pavel Izmailov", "Sharad Vikram", "Matthew D Hoffman", "Andrew Gordon Gordon Wilson" ]
null
null
The posterior over Bayesian neural network (BNN) parameters is extremely high-dimensional and non-convex. For computational reasons, researchers approximate this posterior using inexpensive mini-batch methods such as mean-field variational inference or stochastic-gradient Markov chain Monte Carlo (SGMCMC). To investiga...
[]
null
424
2104.14421
title_snapshot
v139/izzo21a
How to Learn when Data Reacts to Your Model: Performative Gradient Descent
https://proceedings.mlr.press/v139/izzo21a.html
[ "Zachary Izzo", "Lexing Ying", "James Zou" ]
null
null
Performative distribution shift captures the setting where the choice of which ML model is deployed changes the data distribution. For example, a bank which uses the number of open credit lines to determine a customer’s risk of default on a loan may induce customers to open more credit lines in order to improve their c...
[]
null
425
2102.07698
title_snapshot
v139/jaegle21a
Perceiver: General Perception with Iterative Attention
https://proceedings.mlr.press/v139/jaegle21a.html
[ "Andrew Jaegle", "Felix Gimeno", "Andy Brock", "Oriol Vinyals", "Andrew Zisserman", "Joao Carreira" ]
null
null
Biological systems understand the world by simultaneously processing high-dimensional inputs from modalities as diverse as vision, audition, touch, proprioception, etc. The perception models used in deep learning on the other hand are designed for individual modalities, often relying on domain-specific assumptions such...
[]
null
426
2103.03206
title_snapshot
v139/jaegle21b
Imitation by Predicting Observations
https://proceedings.mlr.press/v139/jaegle21b.html
[ "Andrew Jaegle", "Yury Sulsky", "Arun Ahuja", "Jake Bruce", "Rob Fergus", "Greg Wayne" ]
null
null
Imitation learning enables agents to reuse and adapt the hard-won expertise of others, offering a solution to several key challenges in learning behavior. Although it is easy to observe behavior in the real-world, the underlying actions may not be accessible. We present a new method for imitation solely from observatio...
[]
null
427
2107.03851
title_snapshot
v139/jafarov21a
Local Correlation Clustering with Asymmetric Classification Errors
https://proceedings.mlr.press/v139/jafarov21a.html
[ "Jafar Jafarov", "Sanchit Kalhan", "Konstantin Makarychev", "Yury Makarychev" ]
null
null
In the Correlation Clustering problem, we are given a complete weighted graph $G$ with its edges labeled as “similar" and “dissimilar" by a noisy binary classifier. For a clustering $\mathcal{C}$ of graph $G$, a similar edge is in disagreement with $\mathcal{C}$, if its endpoints belong to distinct clusters; and a diss...
[]
null
428
2108.05697
title_snapshot
v139/jagadeesan21a
Alternative Microfoundations for Strategic Classification
https://proceedings.mlr.press/v139/jagadeesan21a.html
[ "Meena Jagadeesan", "Celestine Mendler-Dünner", "Moritz Hardt" ]
null
null
When reasoning about strategic behavior in a machine learning context it is tempting to combine standard microfoundations of rational agents with the statistical decision theory underlying classification. In this work, we argue that a direct combination of these ingredients leads to brittle solution concepts of limited...
[]
null
429
2106.12705
title_snapshot
v139/jain21a
Robust Density Estimation from Batches: The Best Things in Life are (Nearly) Free
https://proceedings.mlr.press/v139/jain21a.html
[ "Ayush Jain", "Alon Orlitsky" ]
null
null
In many applications data are collected in batches, some potentially biased, corrupt, or even adversarial. Learning algorithms for this setting have therefore garnered considerable recent attention. In particular, a sequence of works has shown that all approximately piecewise polynomial distributions—and in particular ...
[]
null
430
null
null
v139/jalal21a
Instance-Optimal Compressed Sensing via Posterior Sampling
https://proceedings.mlr.press/v139/jalal21a.html
[ "Ajil Jalal", "Sushrut Karmalkar", "Alex Dimakis", "Eric Price" ]
null
null
We characterize the measurement complexity of compressed sensing of signals drawn from a known prior distribution, even when the support of the prior is the entire space (rather than, say, sparse vectors). We show for Gaussian measurements and \emph{any} prior distribution on the signal, that the posterior sampling est...
[]
null
431
2106.11438
title_snapshot
v139/jalal21b
Fairness for Image Generation with Uncertain Sensitive Attributes
https://proceedings.mlr.press/v139/jalal21b.html
[ "Ajil Jalal", "Sushrut Karmalkar", "Jessica Hoffmann", "Alex Dimakis", "Eric Price" ]
null
null
This work tackles the issue of fairness in the context of generative procedures, such as image super-resolution, which entail different definitions from the standard classification setting. Moreover, while traditional group fairness definitions are typically defined with respect to specified protected groups – camoufla...
[]
null
432
2106.12182
title_snapshot
v139/jalalzai21a
Feature Clustering for Support Identification in Extreme Regions
https://proceedings.mlr.press/v139/jalalzai21a.html
[ "Hamid Jalalzai", "Rémi Leluc" ]
null
null
Understanding the complex structure of multivariate extremes is a major challenge in various fields from portfolio monitoring and environmental risk management to insurance. In the framework of multivariate Extreme Value Theory, a common characterization of extremes’ dependence structure is the angular measure. It is a...
[]
null
433
2008.07365
title_snapshot
v139/jang21a
Improved Regret Bounds of Bilinear Bandits using Action Space Analysis
https://proceedings.mlr.press/v139/jang21a.html
[ "Kyoungseok Jang", "Kwang-Sung Jun", "Se-Young Yun", "Wanmo Kang" ]
null
null
We consider the bilinear bandit problem where the learner chooses a pair of arms, each from two different action spaces of dimension $d_1$ and $d_2$, respectively. The learner then receives a reward whose expectation is a bilinear function of the two chosen arms with an unknown matrix parameter $\Theta^*\in\mathbb{R}^{...
[]
null
434
null
null
v139/jarrett21a
Inverse Decision Modeling: Learning Interpretable Representations of Behavior
https://proceedings.mlr.press/v139/jarrett21a.html
[ "Daniel Jarrett", "Alihan Hüyük", "Mihaela Van Der Schaar" ]
null
null
Decision analysis deals with modeling and enhancing decision processes. A principal challenge in improving behavior is in obtaining a transparent *description* of existing behavior in the first place. In this paper, we develop an expressive, unifying perspective on *inverse decision modeling*: a framework for learning ...
[]
null
435
2310.18591
title_snapshot
v139/jastrzebski21a
Catastrophic Fisher Explosion: Early Phase Fisher Matrix Impacts Generalization
https://proceedings.mlr.press/v139/jastrzebski21a.html
[ "Stanislaw Jastrzebski", "Devansh Arpit", "Oliver Astrand", "Giancarlo B Kerg", "Huan Wang", "Caiming Xiong", "Richard Socher", "Kyunghyun Cho", "Krzysztof J Geras" ]
null
null
The early phase of training a deep neural network has a dramatic effect on the local curvature of the loss function. For instance, using a small learning rate does not guarantee stable optimization because the optimization trajectory has a tendency to steer towards regions of the loss surface with increasing local curv...
[]
null
436
2012.14193
title_snapshot
v139/javed21a
Policy Gradient Bayesian Robust Optimization for Imitation Learning
https://proceedings.mlr.press/v139/javed21a.html
[ "Zaynah Javed", "Daniel S Brown", "Satvik Sharma", "Jerry Zhu", "Ashwin Balakrishna", "Marek Petrik", "Anca Dragan", "Ken Goldberg" ]
null
null
The difficulty in specifying rewards for many real-world problems has led to an increased focus on learning rewards from human feedback, such as demonstrations. However, there are often many different reward functions that explain the human feedback, leaving agents with uncertainty over what the true reward function is...
[]
null
437
2106.06499
title_snapshot
v139/jayaram21a
In-Database Regression in Input Sparsity Time
https://proceedings.mlr.press/v139/jayaram21a.html
[ "Rajesh Jayaram", "Alireza Samadian", "David Woodruff", "Peng Ye" ]
null
null
Sketching is a powerful dimensionality reduction technique for accelerating algorithms for data analysis. A crucial step in sketching methods is to compute a subspace embedding (SE) for a large matrix $A \in \mathbb{R}^{N \times d}$. SE’s are the primary tool for obtaining extremely efficient solutions for many linear-...
[]
null
438
2107.05672
title_snapshot
v139/jayaram21b
Parallel and Flexible Sampling from Autoregressive Models via Langevin Dynamics
https://proceedings.mlr.press/v139/jayaram21b.html
[ "Vivek Jayaram", "John Thickstun" ]
null
null
This paper introduces an alternative approach to sampling from autoregressive models. Autoregressive models are typically sampled sequentially, according to the transition dynamics defined by the model. Instead, we propose a sampling procedure that initializes a sequence with white noise and follows a Markov chain defi...
[]
null
439
2105.08164
title_snapshot
v139/jeong21a
Objective Bound Conditional Gaussian Process for Bayesian Optimization
https://proceedings.mlr.press/v139/jeong21a.html
[ "Taewon Jeong", "Heeyoung Kim" ]
null
null
A Gaussian process is a standard surrogate model for an unknown objective function in Bayesian optimization. In this paper, we propose a new surrogate model, called the objective bound conditional Gaussian process (OBCGP), to condition a Gaussian process on a bound on the optimal function value. The bound is obtained a...
[]
null
440
null
null
v139/jesson21a
Quantifying Ignorance in Individual-Level Causal-Effect Estimates under Hidden Confounding
https://proceedings.mlr.press/v139/jesson21a.html
[ "Andrew Jesson", "Sören Mindermann", "Yarin Gal", "Uri Shalit" ]
null
null
We study the problem of learning conditional average treatment effects (CATE) from high-dimensional, observational data with unobserved confounders. Unobserved confounders introduce ignorance—a level of unidentifiability—about an individual’s response to treatment by inducing bias in CATE estimates. We present a new pa...
[]
null
441
2103.04850
title_snapshot
v139/jha21a
DeepReDuce: ReLU Reduction for Fast Private Inference
https://proceedings.mlr.press/v139/jha21a.html
[ "Nandan Kumar Jha", "Zahra Ghodsi", "Siddharth Garg", "Brandon Reagen" ]
null
null
The recent rise of privacy concerns has led researchers to devise methods for private neural inference—where inferences are made directly on encrypted data, never seeing inputs. The primary challenge facing private inference is that computing on encrypted data levies an impractically-high latency penalty, stemming most...
[]
null
442
2103.01396
title_snapshot
v139/jha21b
Factor-analytic inverse regression for high-dimension, small-sample dimensionality reduction
https://proceedings.mlr.press/v139/jha21b.html
[ "Aditi Jha", "Michael J. Morais", "Jonathan W Pillow" ]
null
null
Sufficient dimension reduction (SDR) methods are a family of supervised methods for dimensionality reduction that seek to reduce dimensionality while preserving information about a target variable of interest. However, existing SDR methods typically require more observations than the number of dimensions ($N > p$). To ...
[]
null
443
null
null
v139/ji21a
Fast margin maximization via dual acceleration
https://proceedings.mlr.press/v139/ji21a.html
[ "Ziwei Ji", "Nathan Srebro", "Matus Telgarsky" ]
null
null
We present and analyze a momentum-based gradient method for training linear classifiers with an exponentially-tailed loss (e.g., the exponential or logistic loss), which maximizes the classification margin on separable data at a rate of O(1/t^2). This contrasts with a rate of O(1/log(t)) for standard gradient descent, ...
[]
null
444
2107.00595
title_snapshot
v139/ji21b
Marginalized Stochastic Natural Gradients for Black-Box Variational Inference
https://proceedings.mlr.press/v139/ji21b.html
[ "Geng Ji", "Debora Sujono", "Erik B Sudderth" ]
null
null
Black-box variational inference algorithms use stochastic sampling to analyze diverse statistical models, like those expressed in probabilistic programming languages, without model-specific derivations. While the popular score-function estimator computes unbiased gradient estimates, its variance is often unacceptably l...
[]
null
445
null
null
v139/ji21c
Bilevel Optimization: Convergence Analysis and Enhanced Design
https://proceedings.mlr.press/v139/ji21c.html
[ "Kaiyi Ji", "Junjie Yang", "Yingbin Liang" ]
null
null
Bilevel optimization has arisen as a powerful tool for many machine learning problems such as meta-learning, hyperparameter optimization, and reinforcement learning. In this paper, we investigate the nonconvex-strongly-convex bilevel optimization problem. For deterministic bilevel optimization, we provide a comprehensi...
[]
null
446
2010.07962
title_snapshot
v139/jia21a
Efficient Statistical Tests: A Neural Tangent Kernel Approach
https://proceedings.mlr.press/v139/jia21a.html
[ "Sheng Jia", "Ehsan Nezhadarya", "Yuhuai Wu", "Jimmy Ba" ]
null
null
For machine learning models to make reliable predictions in deployment, one needs to ensure the previously unknown test samples need to be sufficiently similar to the training data. The commonly used shift-invariant kernels do not have the compositionality and fail to capture invariances in high-dimensional data in com...
[]
null
447
null
null
v139/jia21b
Scaling Up Visual and Vision-Language Representation Learning With Noisy Text Supervision
https://proceedings.mlr.press/v139/jia21b.html
[ "Chao Jia", "Yinfei Yang", "Ye Xia", "Yi-Ting Chen", "Zarana Parekh", "Hieu Pham", "Quoc Le", "Yun-Hsuan Sung", "Zhen Li", "Tom Duerig" ]
null
null
Pre-trained representations are becoming crucial for many NLP and perception tasks. While representation learning in NLP has transitioned to training on raw text without human annotations, visual and vision-language representations still rely heavily on curated training datasets that are expensive or require expert kno...
[]
null
448
2102.05918
title_snapshot
v139/jia21c
Multi-Dimensional Classification via Sparse Label Encoding
https://proceedings.mlr.press/v139/jia21c.html
[ "Bin-Bin Jia", "Min-Ling Zhang" ]
null
null
In multi-dimensional classification (MDC), there are multiple class variables in the output space with each of them corresponding to one heterogeneous class space. Due to the heterogeneity of class spaces, it is quite challenging to consider the dependencies among class variables when learning from MDC examples. In thi...
[]
null
449
null
null
v139/jiang21a
Self-Damaging Contrastive Learning
https://proceedings.mlr.press/v139/jiang21a.html
[ "Ziyu Jiang", "Tianlong Chen", "Bobak J Mortazavi", "Zhangyang Wang" ]
null
null
The recent breakthrough achieved by contrastive learning accelerates the pace for deploying unsupervised training on real-world data applications. However, unlabeled data in reality is commonly imbalanced and shows a long-tail distribution, and it is unclear how robustly the latest contrastive learning methods could pe...
[]
null
450
2106.02990
title_snapshot
v139/jiang21b
Prioritized Level Replay
https://proceedings.mlr.press/v139/jiang21b.html
[ "Minqi Jiang", "Edward Grefenstette", "Tim Rocktäschel" ]
null
null
Environments with procedurally generated content serve as important benchmarks for testing systematic generalization in deep reinforcement learning. In this setting, each level is an algorithmically created environment instance with a unique configuration of its factors of variation. Training on a prespecified subset o...
[]
null
451
2010.03934
title_snapshot
v139/jiang21c
Monotonic Robust Policy Optimization with Model Discrepancy
https://proceedings.mlr.press/v139/jiang21c.html
[ "Yuankun Jiang", "Chenglin Li", "Wenrui Dai", "Junni Zou", "Hongkai Xiong" ]
null
null
State-of-the-art deep reinforcement learning (DRL) algorithms tend to overfit due to the model discrepancy between source and target environments. Though applying domain randomization during training can improve the average performance by randomly generating a sufficient diversity of environments in simulator, the wors...
[]
null
452
null
null
v139/jiang21d
Approximation Theory of Convolutional Architectures for Time Series Modelling
https://proceedings.mlr.press/v139/jiang21d.html
[ "Haotian Jiang", "Zhong Li", "Qianxiao Li" ]
null
null
We study the approximation properties of convolutional architectures applied to time series modelling, which can be formulated mathematically as a functional approximation problem. In the recurrent setting, recent results reveal an intricate connection between approximation efficiency and memory structures in the data ...
[]
null
453
2107.09355
title_snapshot
v139/jiang21e
Streaming and Distributed Algorithms for Robust Column Subset Selection
https://proceedings.mlr.press/v139/jiang21e.html
[ "Shuli Jiang", "Dennis Li", "Irene Mengze Li", "Arvind V Mahankali", "David Woodruff" ]
null
null
We give the first single-pass streaming algorithm for Column Subset Selection with respect to the entrywise $\ell_p$-norm with $1 \leq p < 2$. We study the $\ell_p$ norm loss since it is often considered more robust to noise than the standard Frobenius norm. Given an input matrix $A \in \mathbb{R}^{d \times n}$ ($n \gg...
[]
null
454
2107.07657
title_snapshot
v139/jiang21f
Single Pass Entrywise-Transformed Low Rank Approximation
https://proceedings.mlr.press/v139/jiang21f.html
[ "Yifei Jiang", "Yi Li", "Yiming Sun", "Jiaxin Wang", "David Woodruff" ]
null
null
In applications such as natural language processing or computer vision, one is given a large $n \times n$ matrix $A = (a_{i,j})$ and would like to compute a matrix decomposition, e.g., a low rank approximation, of a function $f(A) = (f(a_{i,j}))$ applied entrywise to $A$. A very important special case is the likelihood...
[]
null
455
2107.07889
title_snapshot
v139/jiang21g
The Emergence of Individuality
https://proceedings.mlr.press/v139/jiang21g.html
[ "Jiechuan Jiang", "Zongqing Lu" ]
null
null
Individuality is essential in human society. It induces the division of labor and thus improves the efficiency and productivity. Similarly, it should also be a key to multi-agent cooperation. Inspired by that individuality is of being an individual separate from others, we propose a simple yet efficient method for the ...
[]
null
456
2006.05842
title_snapshot
v139/jiang21h
Online Selection Problems against Constrained Adversary
https://proceedings.mlr.press/v139/jiang21h.html
[ "Zhihao Jiang", "Pinyan Lu", "Zhihao Gavin Tang", "Yuhao Zhang" ]
null
null
Inspired by a recent line of work in online algorithms with predictions, we study the constrained adversary model that utilizes predictions from a different perspective. Prior works mostly focused on designing simultaneously robust and consistent algorithms, without making assumptions on the quality of the predictions....
[]
null
457
null
null
v139/jiang21i
Active Covering
https://proceedings.mlr.press/v139/jiang21i.html
[ "Heinrich Jiang", "Afshin Rostamizadeh" ]
null
null
We analyze the problem of active covering, where the learner is given an unlabeled dataset and can sequentially label query examples. The objective is to label query all of the positive examples in the fewest number of total label queries. We show under standard non-parametric assumptions that a classical support estim...
[]
null
458
2106.02552
title_snapshot
v139/jiang21j
Emphatic Algorithms for Deep Reinforcement Learning
https://proceedings.mlr.press/v139/jiang21j.html
[ "Ray Jiang", "Tom Zahavy", "Zhongwen Xu", "Adam White", "Matteo Hessel", "Charles Blundell", "Hado Van Hasselt" ]
null
null
Off-policy learning allows us to learn about possible policies of behavior from experience generated by a different behavior policy. Temporal difference (TD) learning algorithms can become unstable when combined with function approximation and off-policy sampling—this is known as the “deadly triad”. Emphatic temporal d...
[]
null
459
2106.11779
title_snapshot
v139/jiang21k
Characterizing Structural Regularities of Labeled Data in Overparameterized Models
https://proceedings.mlr.press/v139/jiang21k.html
[ "Ziheng Jiang", "Chiyuan Zhang", "Kunal Talwar", "Michael C Mozer" ]
null
null
Humans are accustomed to environments that contain both regularities and exceptions. For example, at most gas stations, one pays prior to pumping, but the occasional rural station does not accept payment in advance. Likewise, deep neural networks can generalize across instances that share common patterns or structures,...
[]
null
460
2002.03206
title_snapshot
v139/jin21a
Optimal Streaming Algorithms for Multi-Armed Bandits
https://proceedings.mlr.press/v139/jin21a.html
[ "Tianyuan Jin", "Keke Huang", "Jing Tang", "Xiaokui Xiao" ]
null
null
This paper studies two variants of the best arm identification (BAI) problem under the streaming model, where we have a stream of n arms with reward distributions supported on [0,1] with unknown means. The arms in the stream are arriving one by one, and the algorithm cannot access an arm unless it is stored in a limite...
[]
null
461
2410.17835
title_snapshot
v139/jin21b
Towards Tight Bounds on the Sample Complexity of Average-reward MDPs
https://proceedings.mlr.press/v139/jin21b.html
[ "Yujia Jin", "Aaron Sidford" ]
null
null
We prove new upper and lower bounds for sample complexity of finding an $\epsilon$-optimal policy of an infinite-horizon average-reward Markov decision process (MDP) given access to a generative model. When the mixing time of the probability transition matrix of all policies is at most $t_\mathrm{mix}$, we provide an a...
[]
null
462
2106.07046
title_snapshot
v139/jin21c
Almost Optimal Anytime Algorithm for Batched Multi-Armed Bandits
https://proceedings.mlr.press/v139/jin21c.html
[ "Tianyuan Jin", "Jing Tang", "Pan Xu", "Keke Huang", "Xiaokui Xiao", "Quanquan Gu" ]
null
null
In batched multi-armed bandit problems, the learner can adaptively pull arms and adjust strategy in batches. In many real applications, not only the regret but also the batch complexity need to be optimized. Existing batched bandit algorithms usually assume that the time horizon T is known in advance. However, many app...
[]
null
463
null
null
v139/jin21d
MOTS: Minimax Optimal Thompson Sampling
https://proceedings.mlr.press/v139/jin21d.html
[ "Tianyuan Jin", "Pan Xu", "Jieming Shi", "Xiaokui Xiao", "Quanquan Gu" ]
null
null
Thompson sampling is one of the most widely used algorithms in many online decision problems due to its simplicity for implementation and superior empirical performance over other state-of-the-art methods. Despite its popularity and empirical success, it has remained an open problem whether Thompson sampling can achiev...
[]
null
464
2003.01803
title_snapshot
v139/jin21e
Is Pessimism Provably Efficient for Offline RL?
https://proceedings.mlr.press/v139/jin21e.html
[ "Ying Jin", "Zhuoran Yang", "Zhaoran Wang" ]
null
null
We study offline reinforcement learning (RL), which aims to learn an optimal policy based on a dataset collected a priori. Due to the lack of further interactions with the environment, offline RL suffers from the insufficient coverage of the dataset, which eludes most existing theoretical analysis. In this paper, we pr...
[]
null
465
2012.15085
title_snapshot
v139/jing21a
Adversarial Option-Aware Hierarchical Imitation Learning
https://proceedings.mlr.press/v139/jing21a.html
[ "Mingxuan Jing", "Wenbing Huang", "Fuchun Sun", "Xiaojian Ma", "Tao Kong", "Chuang Gan", "Lei Li" ]
null
null
It has been a challenge to learning skills for an agent from long-horizon unannotated demonstrations. Existing approaches like Hierarchical Imitation Learning(HIL) are prone to compounding errors or suboptimal solutions. In this paper, we propose Option-GAIL, a novel method to learn skills at long horizon. The key idea...
[]
null
466
2106.05530
title_snapshot
v139/jo21a
Discrete-Valued Latent Preference Matrix Estimation with Graph Side Information
https://proceedings.mlr.press/v139/jo21a.html
[ "Changhun Jo", "Kangwook Lee" ]
null
null
Incorporating graph side information into recommender systems has been widely used to better predict ratings, but relatively few works have focused on theoretical guarantees. Ahn et al. (2018) firstly characterized the optimal sample complexity in the presence of graph side information, but the results are limited due ...
[]
null
467
2003.07040
title_snapshot
v139/jordan21a
Provable Lipschitz Certification for Generative Models
https://proceedings.mlr.press/v139/jordan21a.html
[ "Matt Jordan", "Alex Dimakis" ]
null
null
We present a scalable technique for upper bounding the Lipschitz constant of generative models. We relate this quantity to the maximal norm over the set of attainable vector-Jacobian products of a given generative model. We approximate this set by layerwise convex approximations using zonotopes. Our approach generalize...
[]
null
468
2107.02732
title_snapshot
v139/jorgensen21a
Isometric Gaussian Process Latent Variable Model for Dissimilarity Data
https://proceedings.mlr.press/v139/jorgensen21a.html
[ "Martin Jørgensen", "Soren Hauberg" ]
null
null
We present a probabilistic model where the latent variable respects both the distances and the topology of the modeled data. The model leverages the Riemannian geometry of the generated manifold to endow the latent space with a well-defined stochastic distance measure, which is modeled locally as Nakagami distributions...
[]
null
469
2006.11741
title_snapshot
v139/ju21a
On the Generalization Power of Overfitted Two-Layer Neural Tangent Kernel Models
https://proceedings.mlr.press/v139/ju21a.html
[ "Peizhong Ju", "Xiaojun Lin", "Ness Shroff" ]
null
null
In this paper, we study the generalization performance of min $\ell_2$-norm overfitting solutions for the neural tangent kernel (NTK) model of a two-layer neural network with ReLU activation that has no bias term. We show that, depending on the ground-truth function, the test error of overfitted NTK models exhibits cha...
[]
null
470
2103.05243
title_snapshot
v139/jun21a
Improved Confidence Bounds for the Linear Logistic Model and Applications to Bandits
https://proceedings.mlr.press/v139/jun21a.html
[ "Kwang-Sung Jun", "Lalit Jain", "Blake Mason", "Houssam Nassif" ]
null
null
We propose improved fixed-design confidence bounds for the linear logistic model. Our bounds significantly improve upon the state-of-the-art bound by Li et al. (2017) via recent developments of the self-concordant analysis of the logistic loss (Faury et al., 2020). Specifically, our confidence bound avoids a direct dep...
[]
null
471
2011.11222
title_judge
v139/jung21a
Detection of Signal in the Spiked Rectangular Models
https://proceedings.mlr.press/v139/jung21a.html
[ "Ji Hyung Jung", "Hye Won Chung", "Ji Oon Lee" ]
null
null
We consider the problem of detecting signals in the rank-one signal-plus-noise data matrix models that generalize the spiked Wishart matrices. We show that the principal component analysis can be improved by pre-transforming the matrix entries if the noise is non-Gaussian. As an intermediate step, we prove a sharp phas...
[]
null
472
2104.13517
title_snapshot
v139/jung21b
Estimating Identifiable Causal Effects on Markov Equivalence Class through Double Machine Learning
https://proceedings.mlr.press/v139/jung21b.html
[ "Yonghan Jung", "Jin Tian", "Elias Bareinboim" ]
null
null
General methods have been developed for estimating causal effects from observational data under causal assumptions encoded in the form of a causal graph. Most of this literature assumes that the underlying causal graph is completely specified. However, only observational data is available in most practical settings, wh...
[]
null
473
null
null
v139/kaba21a
A Nullspace Property for Subspace-Preserving Recovery
https://proceedings.mlr.press/v139/kaba21a.html
[ "Mustafa D Kaba", "Chong You", "Daniel P Robinson", "Enrique Mallada", "Rene Vidal" ]
null
null
Much of the theory for classical sparse recovery is based on conditions on the dictionary that are both necessary and sufficient (e.g., nullspace property) or only sufficient (e.g., incoherence and restricted isometry). In contrast, much of the theory for subspace-preserving recovery, the theoretical underpinnings for ...
[]
null
474
null
null
v139/kag21a
Training Recurrent Neural Networks via Forward Propagation Through Time
https://proceedings.mlr.press/v139/kag21a.html
[ "Anil Kag", "Venkatesh Saligrama" ]
null
null
Back-propagation through time (BPTT) has been widely used for training Recurrent Neural Networks (RNNs). BPTT updates RNN parameters on an instance by back-propagating the error in time over the entire sequence length, and as a result, leads to poor trainability due to the well-known gradient explosion/decay phenomena....
[]
null
475
null
null
v139/kairouz21a
The Distributed Discrete Gaussian Mechanism for Federated Learning with Secure Aggregation
https://proceedings.mlr.press/v139/kairouz21a.html
[ "Peter Kairouz", "Ziyu Liu", "Thomas Steinke" ]
null
null
We consider training models on private data that are distributed across user devices. To ensure privacy, we add on-device noise and use secure aggregation so that only the noisy sum is revealed to the server. We present a comprehensive end-to-end system, which appropriately discretizes the data and adds discrete Gaussi...
[]
null
476
2102.06387
title_snapshot
v139/kairouz21b
Practical and Private (Deep) Learning Without Sampling or Shuffling
https://proceedings.mlr.press/v139/kairouz21b.html
[ "Peter Kairouz", "Brendan Mcmahan", "Shuang Song", "Om Thakkar", "Abhradeep Thakurta", "Zheng Xu" ]
null
null
We consider training models with differential privacy (DP) using mini-batch gradients. The existing state-of-the-art, Differentially Private Stochastic Gradient Descent (DP-SGD), requires \emph{privacy amplification by sampling or shuffling} to obtain the best privacy/accuracy/computation trade-offs. Unfortunately, the...
[]
null
477
2103.00039
title_snapshot
v139/kajino21a
A Differentiable Point Process with Its Application to Spiking Neural Networks
https://proceedings.mlr.press/v139/kajino21a.html
[ "Hiroshi Kajino" ]
null
null
This paper is concerned about a learning algorithm for a probabilistic model of spiking neural networks (SNNs). Jimenez Rezende & Gerstner (2014) proposed a stochastic variational inference algorithm to train SNNs with hidden neurons. The algorithm updates the variational distribution using the score function gradient ...
[]
null
478
2106.00901
title_snapshot
v139/kalantzis21a
Projection techniques to update the truncated SVD of evolving matrices with applications
https://proceedings.mlr.press/v139/kalantzis21a.html
[ "Vasileios Kalantzis", "Georgios Kollias", "Shashanka Ubaru", "Athanasios N. Nikolakopoulos", "Lior Horesh", "Kenneth Clarkson" ]
null
null
This submission considers the problem of updating the rank-$k$ truncated Singular Value Decomposition (SVD) of matrices subject to the addition of new rows and/or columns over time. Such matrix problems represent an important computational kernel in applications such as Latent Semantic Indexing and Recommender Systems....
[]
null
479
2010.06392
title_judge
v139/kallus21a
Optimal Off-Policy Evaluation from Multiple Logging Policies
https://proceedings.mlr.press/v139/kallus21a.html
[ "Nathan Kallus", "Yuta Saito", "Masatoshi Uehara" ]
null
null
We study off-policy evaluation (OPE) from multiple logging policies, each generating a dataset of fixed size, i.e., stratified sampling. Previous work noted that in this setting the ordering of the variances of different importance sampling estimators is instance-dependent, which brings up a dilemma as to which importa...
[]
null
480
2010.11002
title_snapshot
v139/kamoutsi21a
Efficient Performance Bounds for Primal-Dual Reinforcement Learning from Demonstrations
https://proceedings.mlr.press/v139/kamoutsi21a.html
[ "Angeliki Kamoutsi", "Goran Banjac", "John Lygeros" ]
null
null
We consider large-scale Markov decision processes with an unknown cost function and address the problem of learning a policy from a finite set of expert demonstrations. We assume that the learner is not allowed to interact with the expert and has no access to reinforcement signal of any kind. Existing inverse rei...
[]
null
481
2112.14004
title_snapshot
v139/kandiros21a
Statistical Estimation from Dependent Data
https://proceedings.mlr.press/v139/kandiros21a.html
[ "Vardis Kandiros", "Yuval Dagan", "Nishanth Dikkala", "Surbhi Goel", "Constantinos Daskalakis" ]
null
null
We consider a general statistical estimation problem wherein binary labels across different observations are not independent conditioning on their feature vectors, but dependent, capturing settings where e.g. these observations are collected on a spatial domain, a temporal domain, or a social network, which induce depe...
[]
null
482
2107.09773
title_snapshot
v139/kapoor21a
SKIing on Simplices: Kernel Interpolation on the Permutohedral Lattice for Scalable Gaussian Processes
https://proceedings.mlr.press/v139/kapoor21a.html
[ "Sanyam Kapoor", "Marc Finzi", "Ke Alexander Wang", "Andrew Gordon Gordon Wilson" ]
null
null
State-of-the-art methods for scalable Gaussian processes use iterative algorithms, requiring fast matrix vector multiplies (MVMs) with the co-variance kernel. The Structured Kernel Interpolation (SKI) framework accelerates these MVMs by performing efficient MVMs on a grid and interpolating back to the original space. I...
[]
null
483
2106.06695
title_snapshot
v139/kapoor21b
Variational Auto-Regressive Gaussian Processes for Continual Learning
https://proceedings.mlr.press/v139/kapoor21b.html
[ "Sanyam Kapoor", "Theofanis Karaletsos", "Thang D Bui" ]
null
null
Through sequential construction of posteriors on observing data online, Bayes’ theorem provides a natural framework for continual learning. We develop Variational Auto-Regressive Gaussian Processes (VAR-GPs), a principled posterior updating mechanism to solve sequential tasks in continual learning. By relying on sparse...
[]
null
484
2006.05468
title_snapshot
v139/karampatziakis21a
Off-Policy Confidence Sequences
https://proceedings.mlr.press/v139/karampatziakis21a.html
[ "Nikos Karampatziakis", "Paul Mineiro", "Aaditya Ramdas" ]
null
null
We develop confidence bounds that hold uniformly over time for off-policy evaluation in the contextual bandit setting. These confidence sequences are based on recent ideas from martingale analysis and are non-asymptotic, non-parametric, and valid at arbitrary stopping times. We provide algorithms for computing these co...
[]
null
485
2102.09540
title_snapshot
v139/karimireddy21a
Learning from History for Byzantine Robust Optimization
https://proceedings.mlr.press/v139/karimireddy21a.html
[ "Sai Praneeth Karimireddy", "Lie He", "Martin Jaggi" ]
null
null
Byzantine robustness has received significant attention recently given its importance for distributed and federated learning. In spite of this, we identify severe flaws in existing algorithms even when the data across the participants is identically distributed. First, we show realistic examples where current state of ...
[]
null
486
2012.10333
title_snapshot
v139/kato21a
Non-Negative Bregman Divergence Minimization for Deep Direct Density Ratio Estimation
https://proceedings.mlr.press/v139/kato21a.html
[ "Masahiro Kato", "Takeshi Teshima" ]
null
null
Density ratio estimation (DRE) is at the core of various machine learning tasks such as anomaly detection and domain adaptation. In the DRE literature, existing studies have extensively studied methods based on Bregman divergence (BD) minimization. However, when we apply the BD minimization with highly flexible models,...
[]
null
487
2006.06979
title_snapshot
v139/katz-samuels21a
Improved Algorithms for Agnostic Pool-based Active Classification
https://proceedings.mlr.press/v139/katz-samuels21a.html
[ "Julian Katz-Samuels", "Jifan Zhang", "Lalit Jain", "Kevin Jamieson" ]
null
null
We consider active learning for binary classification in the agnostic pool-based setting. The vast majority of works in active learning in the agnostic setting are inspired by the CAL algorithm where each query is uniformly sampled from the disagreement region of the current version space. The sample complexity of such...
[]
null
488
2105.06499
title_snapshot
v139/kaya21a
When Does Data Augmentation Help With Membership Inference Attacks?
https://proceedings.mlr.press/v139/kaya21a.html
[ "Yigitcan Kaya", "Tudor Dumitras" ]
null
null
Deep learning models often raise privacy concerns as they leak information about their training data. This leakage enables membership inference attacks (MIA) that can identify whether a data point was in a model’s training set. Research shows that some ’data augmentation’ mechanisms may reduce the risk by combatting a ...
[]
null
489
null
null
v139/kazemi21a
Regularized Submodular Maximization at Scale
https://proceedings.mlr.press/v139/kazemi21a.html
[ "Ehsan Kazemi", "Shervin Minaee", "Moran Feldman", "Amin Karbasi" ]
null
null
In this paper, we propose scalable methods for maximizing a regularized submodular function $f \triangleq g-\ell$ expressed as the difference between a monotone submodular function $g$ and a modular function $\ell$. Submodularity is inherently related to the notions of diversity, coverage, and representativeness. In pa...
[]
null
490
2002.03503
title_snapshot
v139/kelkar21a
Prior Image-Constrained Reconstruction using Style-Based Generative Models
https://proceedings.mlr.press/v139/kelkar21a.html
[ "Varun A Kelkar", "Mark Anastasio" ]
null
null
Obtaining a useful estimate of an object from highly incomplete imaging measurements remains a holy grail of imaging science. Deep learning methods have shown promise in learning object priors or constraints to improve the conditioning of an ill-posed imaging inverse problem. In this study, a framework for estimating a...
[]
null
491
2102.12525
title_snapshot
v139/keller21a
Self Normalizing Flows
https://proceedings.mlr.press/v139/keller21a.html
[ "Thomas A Keller", "Jorn W.T. Peters", "Priyank Jaini", "Emiel Hoogeboom", "Patrick Forré", "Max Welling" ]
null
null
Efficient gradient computation of the Jacobian determinant term is a core problem in many machine learning settings, and especially so in the normalizing flow framework. Most proposed flow models therefore either restrict to a function class with easy evaluation of the Jacobian determinant, or an efficient estimator th...
[]
null
492
2011.07248
title_snapshot
v139/kenlay21a
Interpretable Stability Bounds for Spectral Graph Filters
https://proceedings.mlr.press/v139/kenlay21a.html
[ "Henry Kenlay", "Dorina Thanou", "Xiaowen Dong" ]
null
null
Graph-structured data arise in a variety of real-world context ranging from sensor and transportation to biological and social networks. As a ubiquitous tool to process graph-structured data, spectral graph filters have been used to solve common tasks such as denoising and anomaly detection, as well as design deep lear...
[]
null
493
2102.09587
title_snapshot
v139/kerdreux21a
Affine Invariant Analysis of Frank-Wolfe on Strongly Convex Sets
https://proceedings.mlr.press/v139/kerdreux21a.html
[ "Thomas Kerdreux", "Lewis Liu", "Simon Lacoste-Julien", "Damien Scieur" ]
null
null
It is known that the Frank-Wolfe (FW) algorithm, which is affine covariant, enjoys faster convergence rates than $\mathcal{O}\left(1/K\right)$ when the constraint set is strongly convex. However, these results rely on norm-dependent assumptions, usually incurring non-affine invariant bounds, in contradiction with FW’s ...
[]
null
494
2011.03351
title_snapshot
v139/khachaturov21a
Markpainting: Adversarial Machine Learning meets Inpainting
https://proceedings.mlr.press/v139/khachaturov21a.html
[ "David Khachaturov", "Ilia Shumailov", "Yiren Zhao", "Nicolas Papernot", "Ross Anderson" ]
null
null
Inpainting is a learned interpolation technique that is based on generative modeling and used to populate masked or missing pieces in an image; it has wide applications in picture editing and retouching. Recently, inpainting started being used for watermark removal, raising concerns. In this paper we study how to manip...
[]
null
495
2106.00660
title_snapshot
v139/khodadadian21a
Finite-Sample Analysis of Off-Policy Natural Actor-Critic Algorithm
https://proceedings.mlr.press/v139/khodadadian21a.html
[ "Sajad Khodadadian", "Zaiwei Chen", "Siva Theja Maguluri" ]
null
null
In this paper, we provide finite-sample convergence guarantees for an off-policy variant of the natural actor-critic (NAC) algorithm based on Importance Sampling. In particular, we show that the algorithm converges to a global optimal policy with a sample complexity of $\mathcal{O}(\epsilon^{-3}\log^2(1/\epsilon))$ und...
[]
null
496
2102.09318
title_snapshot
v139/khrulkov21a
Functional Space Analysis of Local GAN Convergence
https://proceedings.mlr.press/v139/khrulkov21a.html
[ "Valentin Khrulkov", "Artem Babenko", "Ivan Oseledets" ]
null
null
Recent work demonstrated the benefits of studying continuous-time dynamics governing the GAN training. However, this dynamics is analyzed in the model parameter space, which results in finite-dimensional dynamical systems. We propose a novel perspective where we study the local dynamics of adversarial training in the g...
[]
null
497
2102.04448
title_snapshot
v139/kidger21a
"Hey, that’s not an ODE": Faster ODE Adjoints via Seminorms
https://proceedings.mlr.press/v139/kidger21a.html
[ "Patrick Kidger", "Ricky T. Q. Chen", "Terry J Lyons" ]
null
null
Neural differential equations may be trained by backpropagating gradients via the adjoint method, which is another differential equation typically solved using an adaptive-step-size numerical differential equation solver. A proposed step is accepted if its error, \emph{relative to some norm}, is sufficiently small; els...
[]
null
498
2009.09457
title_snapshot
v139/kidger21b
Neural SDEs as Infinite-Dimensional GANs
https://proceedings.mlr.press/v139/kidger21b.html
[ "Patrick Kidger", "James Foster", "Xuechen Li", "Terry J Lyons" ]
null
null
Stochastic differential equations (SDEs) are a staple of mathematical modelling of temporal dynamics. However, a fundamental limitation has been that such models have typically been relatively inflexible, which recent work introducing Neural SDEs has sought to solve. Here, we show that the current classical approach to...
[]
null
499
2102.03657
title_snapshot
v139/killamsetty21a
GRAD-MATCH: Gradient Matching based Data Subset Selection for Efficient Deep Model Training
https://proceedings.mlr.press/v139/killamsetty21a.html
[ "Krishnateja Killamsetty", "Durga S", "Ganesh Ramakrishnan", "Abir De", "Rishabh Iyer" ]
null
null
The great success of modern machine learning models on large datasets is contingent on extensive computational resources with high financial and environmental costs. One way to address this is by extracting subsets that generalize on par with the full data. In this work, we propose a general framework, GRAD-MATCH, whic...
[]
null
500
2103.00123
title_snapshot