arXiv research

Proposes a transductive matrix completion method with calibration for multi-task learning.

Proposes SPFB method for optimizing partition functions in stochastic learning.

In this work, we study the problem of aggregating a finite number of predictors for nonstationary sub-linear processes. We provide oracle inequalities relying essentially on three ingredients: (1) a uniform bound of the $\ell^1$ norm of the time varying sub-linear coefficients, (2) a Lipschitz assumption on the predict…

New BO method optimizes functions efficiently even with unknown hyperparameters.

Policy gradient methods achieve linear convergence in simple MDPs.

Efficient algorithms for $k$-means clustering frequently converge to suboptimal partitions, and given a partition, it is difficult to detect $k$-means optimality. In this paper, we develop an a posteriori certifier of approximate optimality for $k$-means clustering. The certifier is a sub-linear Monte Carlo algorithm b…

Bayesian optimisation algorithm for unknown search spaces with sub-linear regret.

New method speeds up diffusion models inference to sub-linear time.

Study finds exact limits for sparse regression with fewer observations than usual.

Kernel $ε$-Greedy optimizes multi-armed bandits with covariates for sub-linear regret.

Gradient EM converges globally for over-parameterized Gaussian mixtures.

New model for multi-armed bandits with growing arms.

We propose the stochastic average gradient (SAG) method for optimizing the sum of a finite number of smooth convex functions. Like stochastic gradient (SG) methods, the SAG method's iteration cost is independent of the number of terms in the sum. However, by incorporating a memory of previous gradient values the SAG me…

We revisit the inductive matrix completion problem that aims to recover a rank-$r$ matrix with ambient dimension $d$ given $n$ features as the side prior information. The goal is to make use of the known $n$ features to reduce sample and computational complexities. We present and analyze a new gradient-based non-convex…

In this paper, we study a class of online optimization problems with long-term budget constraints where the objective functions are not necessarily concave (nor convex) but they instead satisfy the Diminishing Returns (DR) property. Specifically, a sequence of monotone DR-submodular objective functions $\{f_t(x)\}_{t=1…

The paper analyzes mirror descent in measure spaces and its applications.

Optimal Liouville theorem for minimal disks in any codimension.

Study designs steering rewards for MFGs with unknown dynamics and model uncertainty.

Paper tackles LDP bandits learning with improved results and sub-linear regret.

First robust bandit algorithm for contextual bandits with sub-linear regret.

New algorithm tackles adversarial corruption in Lipschitz bandits with sub-linear regret.

LNUCB-TA improves MAB performance by dynamically adjusting exploration rates and recognizing spatiotemporal patterns.

In this paper, we study the problem of constrained robust (min-max) optimization ina black-box setting, where the desired optimizer cannot access the gradients of the objective function but may query its values. We present a principled optimization framework, integrating a zeroth-order (ZO) gradient estimator with an a…

Polyak-Ruppert CLT for SA-Adam with momentum and non-convergent adaptive preconditioning

Study online learning with individual fairness without known similarity measure.

The multinomial logistic regression (MLR) model is widely used in statistics and machine learning. Stochastic gradient descent (SGD) is the most common approach for determining the parameters of a MLR model in big data scenarios. However, SGD has slow sub-linear rates of convergence. A way to improve these rates of con…

Distributed data-parallel algorithms aim to accelerate the training of deep neural networks by parallelizing the computation of large mini-batch gradient updates across multiple nodes. Approaches that synchronize nodes using exact distributed averaging (e.g., via AllReduce) are sensitive to stragglers and communication…

In this paper we want to exploit further the semi-discrete method appeared in Halidias and Stamatiou (2015). We are interested in the numerical solution of mean reverting CEV processes that appear in financial mathematics models and are described as non negative solutions of certain stochastic differential equations wi…

DynBRO learns robustly from dynamic Byzantine workers.

In $\mathcal{X}$-armed bandit problem an agent sequentially interacts with environment which yields a reward based on the vector input the agent provides. The agent's goal is to maximise the sum of these rewards across some number of time steps. The problem and its variations have been a subject of numerous studies, su…

In this work and the supporting Part II, we examine the performance of stochastic sub-gradient learning strategies under weaker conditions than usually considered in the literature. The new conditions are shown to be automatically satisfied by several important cases of interest including SVM, LASSO, and Total-Variatio…

Bayesian optimization guided by experimenter intuition and beliefs.

New algorithms reduce slate bandit regret for large slates, outperforming existing methods.

Deep Retrieval learns a retrievable structure for efficient large-scale recommendations.

Adaptive LR improves neural network Lipschitz regularity without slowing convergence.

Stochastic gradient algorithms estimate the gradient based on only one or a few samples and enjoy low computational cost per iteration. They have been widely used in large-scale optimization problems. However, stochastic gradient algorithms are usually slow to converge and achieve sub-linear convergence rates, due to t…

In this paper, we consider the problem of preserving privacy in the online learning setting. We study the problem in the online convex programming (OCP) framework---a popular online learning setting with several interesting theoretical and practical implications---while using differential privacy as the formal privacy …

In recent years there has been an increasing interest in learning Bayesian networks from data. One of the most effective methods for learning such networks is based on the minimum description length (MDL) principle. Previous work has shown that this learning procedure is asymptotically successful: with probability one,…

New algorithms for planning with adversarial changes in costs.

Paper tackles domain adaptation for contextual bandits with sub-linear regret.

Study of deep Stable neural networks with various activation functions.

New algorithms improve SGD convergence and reduce variance for over-parameterized models.

COBRA addresses strategic behavior in online platforms by ensuring truthful reporting without monetary incentives.

New algorithm for reinforcement learning reduces complexity and guarantees convergence.

Safe Gaussian Process Bandit Optimization with sub-linear regret bounds.

We provide a novel -- and to the best of our knowledge, the first -- algorithm for high dimensional sparse regression with constant fraction of corruptions in explanatory and/or response variables. Our algorithm recovers the true sparse parameters with sub-linear sample complexity, in the presence of a constant fractio…

New algorithm reduces risk in online games with limited feedback.

Upper bounds on ribbonlength of various knots, showing linear and sub-linear behavior.