arXiv research

New algorithms sample convex bodies using Markov chains and restricted Gaussian oracles.

We consider a novel formulation of the multi-armed bandit model, which we call the contextual bandit with restricted context, where only a limited number of features can be accessed by the learner at every iteration. This novel formulation is motivated by different online problems arising in clinical trials, recommende…

Current neural network-based classifiers are susceptible to adversarial examples even in the black-box setting, where the attacker only has query access to the model. In practice, the threat model for real-world systems is often more restrictive than the typical black-box model where the adversary can observe the full …

MixDiff detects OOD samples in constrained access environments by comparing perturbed samples.

We give algorithms for estimating the expectation of a given real-valued function $φ:X\to {\bf R}$ on a sample drawn randomly from some unknown distribution $D$ over domain $X$, namely ${\bf E}_{{\bf x}\sim D}[φ({\bf x})]$. Our algorithms work in two well-studied models of restricted access to data samples. The first o…

Generative Distributionally Robust Optimization (GDRO) improves model compatibility and adversarial structure in DRO.

Federated Learning tackles limited user participation with a new risk-aware approach.

Note that this paper is superceded by "Black-Box Adversarial Attacks with Limited Queries and Information." Current neural network-based image classifiers are susceptible to adversarial examples, even in the black-box setting, where the attacker is limited to query access without access to gradients. Previous methods -…

Uniform sampling of modest size is a coreset for regularized loss minimization.

Study creates open-access wildfire dataset for Russia.

Minimalistic attacks reveal deep RL policies' vulnerabilities with little perturbation.

Data de-duplication is the task of detecting multiple records that correspond to the same real-world entity in a database. In this work, we view de-duplication as a clustering problem where the goal is to put records corresponding to the same physical entity in the same cluster and putting records corresponding to diff…

Proposes a framework for balancing fairness and accuracy in data-restricted binary classification.

Federated learning improves bioinformatics by sharing data legally.

Paper analyzes iterative learning for concept classes and learns half-spaces.

NTL protects AI models by restricting their generalization ability to specific domains.

We develop a parallel variational inference (VI) procedure for use in data-distributed settings, where each machine only has access to a subset of data and runs VI independently, without communicating with other machines. This type of "embarrassingly parallel" procedure has recently been developed for MCMC inference al…

Paper presents a novel time series clustering algorithm for financial inclusion.

RelEx explains relational models without gradient access.

The presence of data corruption in user-generated streaming data, such as social media, motivates a new fundamental problem that learns reliable regression coefficient when features are not accessible entirely at one time. Until now, several important challenges still cannot be handled concurrently: 1) corrupted data e…

We consider the problem of online planning in a Markov Decision Process when given only access to a generative model, restricted to open-loop policies - i.e. sequences of actions - and under budget constraint. In this setting, the Open-Loop Optimistic Planning (OLOP) algorithm enjoys good theoretical guarantees but is …

Suppose an agent is in a (possibly unknown) Markov Decision Process in the absence of a reward signal, what might we hope that an agent can efficiently learn to do? This work studies a broad class of objectives that are defined solely as functions of the state-visitation frequencies that are induced by how the agent be…

With rapid development of the Internet, web contents become huge. Most of the websites are publicly available, and anyone can access the contents from anywhere such as workplace, home and even schools. Nevertheless, not all the web contents are appropriate for all users, especially children. An example of these content…

New method reduces variance and bias in approximating indefinite kernels.

Recent advances in cryptography promise to enable secure statistical computation on encrypted data, whereby a limited set of operations can be carried out without the need to first decrypt. We review these homomorphic encryption schemes in a manner accessible to statisticians and machine learners, focusing on pertinent…

Paper develops methods for fair insurance pricing without direct access to sensitive attributes.

A cost-effective framework for gradual domain adaptation using multifidelity.

WAFFLE embeds watermarks in federated learning models without access to training data.

EnKG solves inverse problems without derivatives, using diffusion models.

Proposes a method to enforce fairness in machine learning models without sensitive data.

Algorithm samples from composite log-concave distributions using gradient evaluations and restricted Gaussian oracles.

By restricting the iterate on a nonlinear manifold, the recently proposed Riemannian optimization methods prove to be both efficient and effective in low rank tensor completion problems. However, existing methods fail to exploit the easily accessible side information, due to their format mismatch. Consequently, there i…

This paper is an updated version of a survey on projective configurations of subspaces in general position. The preceding version was published in Russian in 1989 and in English in 1990 (in Leningrad Math. J.) opening a new section ``Light reading for the professional''. The paper is written in the form of introduction…

Inverse function theorem and homotopy description for L-infinity bundles.

Scalable algorithm for computing Wasserstein-2 barycenters without bias.

Paper discusses ethical norms for machine learning to prevent misuse.

Motivated by settings in which predictive models may be required to be non-discriminatory with respect to certain attributes (such as race), but even collecting the sensitive attribute may be forbidden or restricted, we initiate the study of fair learning under the constraint of differential privacy. We design two lear…

New method improves decision-making accuracy without complex calculations.

We study the minimization of a convex function $f(X)$ over the set of $n\times n$ positive semi-definite matrices, but when the problem is recast as $\min_U g(U) := f(UU^\top)$, with $U \in \mathbb{R}^{n \times r}$ and $r \leq n$. We study the performance of gradient descent on $g$---which we refer to as Factored Gradi…

Simplified Khovanov-Rozansky calculus for bipartite knots.

Recently the authors have explored new concepts of plurisubharmonicity and pseudoconvexity, with much of the attendant analysis, in the context of calibrated manifolds. Here a much broader extension is made. This development covers a wide variety of geometric situations, including, for example, Lagrangian plurisubhamon…

Lower bounds on queries needed for finding stationary points in non-convex optimization.

New algorithm approximates continuous Wasserstein barycenters efficiently.

KT models improved slightly with synthetic student data.

In order to reduce signalling, traders may resort to limiting access to dark venues and imposing limits on minimum fill sizes they are willing to trade. However, doing this also restricts the liquidity available to the trader since an ever increasing quantity of orders are traded by algos in clips. An alternative is to…

Quantile regression attacks outperform shadow models in unseen class membership inference attacks.

New method for estimating local structure around target nodes in DAGs.

High-dimensional settings, where the data dimension ($d$) far exceeds the number of observations ($n$), are common in many statistical and machine learning applications. Methods based on $\ell_1$-relaxation, such as Lasso, are very popular for sparse recovery in these settings. Restricted Eigenvalue (RE) condition is a…