Having similar behavior at training time and test time $-$ what we call a "What You See Is What You Get" (WYSIWYG) property $-$ is desirable in machine learning. Models trained with standard stochastic gradient descent (SGD), however, do not necessarily have this property, as their complex behaviors such as robustness or subgroup performance can differ drastically between training and test time. In contrast, we show that Differentially-Private (DP) training provably ensures the high-level WYSIWYG property, which we quantify using a notion of distributional generalization. Applying this connection, we introduce new conceptual tools for designing deep-learning methods by reducing generalization concerns to optimization ones: to mitigate unwanted behavior at test time, it is provably sufficient to mitigate this behavior on the training data. By applying this novel design principle, which bypasses "pathologies" of SGD, we construct simple algorithms that are competitive with SOTA in several distributional-robustness applications, significantly improve the privacy vs. disparate impact trade-off of DP-SGD, and mitigate robust overfitting in adversarial training. Finally, we also improve on theoretical bounds relating DP, stability, and distributional generalization.
Stanford University（斯坦福大学）；Istituto Italiano di Tecnologia（意大利理工学院）
Neural networks are vulnerable to adversarial examples and researchers have proposed many heuristic attack and defense mechanisms. We address this problem through the principled lens of distributionally robust optimization, which guarantees performance under adversarial input perturbations. By considering a Lagrangian penalty formulation of perturbing the underlying data distribution in a Wasserstein ball, we provide a training procedure that augments model parameter updates with worst-case perturbations of training data. For smooth losses, our procedure provably achieves moderate levels of ro…
Image hashing is a principled approximate nearest neighbor approach to find similar items to a query in a large collection of images. Hashing aims to learn a binary-output function that maps an image to a binary vector. For optimal retrieval performance, producing balanced hash codes with low-quantization error to bridge the gap between the learning stage's continuous relaxation and the inference stage's discrete quantization is important. However, in the existing deep supervised hashing methods, coding balance and low-quantization error are difficult to achieve and involve several losses. We …
Robustness to distributional shift is one of the key challenges of contemporary machine learning. Attaining such robustness is the goal of distributionally robust optimization, which seeks a solution to an optimization problem that is worst-case robust under a specified distributional shift of an uncontrolled covariate. In this paper, we study such a problem when the distributional shift is measured via the maximum mean discrepancy (MMD). For the setting of zeroth-order, noisy optimization, we present a novel distributionally robust Bayesian optimization algorithm (DRBO). Our algorithm provabl…
Learning Credal Ensembles via Distributionally Robust Optimization
Department of Computer Science, KU Leuven, Belgium（比利时库勒万大学计算机科学系）；College of Computing and Data Science, Nanyang Technological University, Singapore（新加坡南洋理工大学计算与数据科学学院）；Department of Mechanical Engineering, KU Leuven, Belgium（比利时库勒万大学机械工程系）；School of Engineering, Computing and Mathematics, Oxford Brookes University, U.K.（英国奥德赛布鲁克斯大学工程、计算与数学学院）
Existing measures and representations for trajectories have two longstanding fundamental shortcomings, i.e., they are computationally expensive and they can not guarantee the `uniqueness' property of a distance function: dist(X,Y) = 0 if and only if X=Y, where $X$ and $Y$ are two trajectories. This paper proposes a simple yet powerful way to represent trajectories and measure the similarity between two trajectories using a distributional kernel to address these shortcomings. It is a principled approach based on kernel mean embedding which has a strong theoretical underpinning. It has three dis…
IBM Research AI（IBM研究院人工智能中心）；Colorado State University-Pueblo（科罗拉多州立大学普韦布洛分校）
Distributionally robust optimization (DRO) problems are increasingly seen as a viable method to train machine learning models for improved model generalization. These min-max formulations, however, are more difficult to solve. We therefore provide a new stochastic gradient descent algorithm to efficiently solve this DRO formulation. Our approach applies gradient descent to the outer minimization formulation and estimates the gradient of the inner maximization based on a sample average approximation. The latter uses a subset of the data in each iteration, progressively increasing the subset siz…
We propose new, more efficient targeted white-box attacks against deep neural networks. Our attacks better align with the attacker's goal: (1) tricking a model to assign higher probability to the target class than to any other class, while (2) staying within an $ε$-distance of the attacked input. First, we demonstrate a loss function that explicitly encodes (1) and show that Auto-PGD finds more attacks with it. Second, we propose a new attack method, Constrained Gradient Descent (CGD), using a refinement of our loss function that captures both (1) and (2). CGD seeks to satisfy both attacker ob…
NTT Communication Science Laboratories（NTT通信科学实验室）；Kyoto University（京都大学）
Finding the features relevant to the difference in treatment effects is essential to unveil the underlying causal mechanisms. Existing methods seek such features by measuring how greatly the feature attributes affect the degree of the {\it conditional average treatment effect} (CATE). However, these methods may overlook important features because CATE, a measure of the average treatment effect, cannot detect differences in distribution parameters other than the mean (e.g., variance). To resolve this weakness of existing methods, we propose a feature selection framework for discovering {\it dis…
University of Wisconsin-Madison（威斯康星大学麦迪逊分校）；American Family Insurance（美国家庭保险集团）
With the rise of big data analytics, multi-layer neural networks have surfaced as one of the most powerful machine learning methods. However, their theoretical mathematical properties are still not fully understood. Training a neural network requires optimizing a non-convex objective function, typically done using stochastic gradient descent (SGD). In this paper, we seek to extend the mean field results of Mei et al. (2018) from two-layer neural networks with one hidden layer to three-layer neural networks with two hidden layers. We will show that the SGD dynamics is captured by a set of non-l…
Distributional Matrix Completion via Nearest Neighbors in the Wasserstein Space
Industrial Engineering and Operations Research, Columbia University（哥伦比亚大学工业工程与运筹学系）；Statistics and Data Science, Cornell Tech, Cornell University（康奈尔大学数据科学与统计学系）；Operations Research and Information Engineering, Cornell Tech, Cornell University（康奈尔大学运筹学与信息工程系）
Current methodologies in machine learning analyze the effects of various statistical parity notions of fairness primarily in light of their impacts on predictive accuracy and vendor utility loss. In this paper, we propose a new framework for interpreting the effects of fairness criteria by converting the constrained loss minimization problem into a social welfare maximization problem. This translation moves a classifier and its output into utility space where individuals, groups, and society at-large experience different welfare changes due to classification assignments. Under this characteriz…
Reinforcement learning (RL) has achieved remarkable performance in numerous sequential decision making and control tasks. However, a common problem is that learned nearly optimal policy always overfits to the training environment and may not be extended to situations never encountered during training. For practical applications, the randomness of environment usually leads to some devastating events, which should be the focus of safety-critical systems such as autonomous driving. In this paper, we introduce the minimax formulation and distributional framework to improve the generalization abili…
It is well-known that deep neural networks (DNNs) are susceptible to adversarial attacks, exposing a severe fragility of deep learning systems. As the result, adversarial training (AT) method, by incorporating adversarial examples during training, represents a natural and effective approach to strengthen the robustness of a DNN-based classifier. However, most AT-based methods, notably PGD-AT and TRADES, typically seek a pointwise adversary that generates the worst-case adversarial example by independently perturbing each data sample, as a way to "probe" the vulnerability of the classifier. Arg…
On Generalization and Distributional Update for Mimicking Observations with Adequate Exploration
Shanghai Institute of Satellite Engineering（上海卫星工程研究所）；Department of Mathematics, College of Sciences, Shanghai University（上海大学数学系）；School of Computer Science, University of Bristol（布里斯托大学计算机科学学院）；Newtouch Center for Mathematics of Shanghai University（上海大学数学中心）
Arizona State University（亚利桑那州立大学）；University of Illinois Chicago（伊利诺伊大学芝加哥分校）；Bytedance AI Lab（字节跳动人工智能实验室）
Machine learning algorithms typically assume that the training and test samples come from the same distributions, i.e., in-distribution. However, in open-world scenarios, streaming big data can be Out-Of-Distribution (OOD), rendering these algorithms ineffective. Prior solutions to the OOD challenge seek to identify invariant features across different training domains. The underlying assumption is that these invariant features should also work reasonably well in the unlabeled target domain. By contrast, this work is interested in the domain-specific features that include both invariant feature…
Several language applications often require word semantics as a core part of their processing pipeline, either as precise meaning inference or semantic similarity. Multi-sense embeddings (M-SE) can be exploited for this important requirement. M-SE seeks to represent each word by their distinct senses in order to resolve the conflation of meanings of words as used in different contexts. Previous works usually approach this task by training a model on a large corpus and often ignore the effect and usefulness of the semantic relations offered by lexical resources. However, even with large trainin…
BlinDNO: A Distributional Neural Operator for Dynamical System Reconstruction from Time-Label-Free data
Department of Mathematical Sciences, Tsinghua University, Beijing, China（清华大学数学科学系）；Yau Mathematical Sciences Center, Tsinghua University, Beijing, China（清华大学姚期智信息科学研究院）；Yanqi Lake Beijing Institute of Mathematical Sciences and Applications, Beijing, China（北京燕琦湖数学科学与应用研究院）
