Publications de Ya-Ping Hsieh | EPFL Graph Search

No-regret learning in games with noisy feedback: Faster rates and adaptivity via learning rate separation

Volkan Cevher, Kimon Antonakopoulos, Ya-Ping Hsieh

We examine the problem of regret minimization when the learner is involved in a continuous game with other optimizing agents: in this case, if all players follow a no-regret algorithm, it is possible to achieve significantly lower regret relative to fully ...

2022

Conditional gradient methods for stochastically constrained convex minimization

Volkan Cevher, Ya-Ping Hsieh, Ahmet Alacaoglu, Maria-Luiza Vladarean

We propose two novel conditional gradient-based methods for solving structured stochastic convex optimization problems with a large number of linear constraints. Instances of this template naturally arise from SDP-relaxations of combinatorial problems, whi ...

2020

Convergence without Convexity: Sampling, Optimization, and Games

Ya-Ping Hsieh

Many important problems in contemporary machine learning involve solving highly non- convex problems in sampling, optimization, or games. The absence of convexity poses significant challenges to convergence analysis of most training algorithms, and in some ...

EPFL2020

Robust Reinforcement Learning via Adversarial training with Langevin Dynamics

Volkan Cevher, Paul Thierry Yves Rolland, Ya-Ping Hsieh, Kamalaruban Parameswaran, Yu-Ting Huang, Cheng Shi

We introduce a sampling perspective to tackle the challenging task of training robust Reinforcement Learning (RL) agents. Leveraging the powerful Stochastic Gradient Langevin Dynamics, we present a novel, scalable two-player RL algorithm, which is a sampli ...

34th Conference on Neural Information Processing Systems (NeurIPS 2020)2020

Dimension-free Information Concentration via Exp-Concavity

Volkan Cevher, Ya-Ping Hsieh

Information concentration of probability measures have important implications in learning theory. Recently, it is discovered that the information content of a log-concave distribution concentrates around their differential entropy, albeit with an unpleasan ...

2018

Let’s be honest: An optimal no-regret framework for zero-sum games

Volkan Cevher, Mehmet Fatih Sahin, Ya-Ping Hsieh, Ehsan Asadi Kangarshahi

We revisit the problem of solving two-player zero- sum games in the decentralized setting. We pro- pose a simple algorithmic framework that simulta- neously achieves the best rates for honest regret as well as adversarial regret, and in addition resolves t ...

2018

A Non-Euclidean Gradient Descent Framework for Non-Convex Matrix Factorization

Volkan Cevher, Anastasios Kyrillidis, Ya-Ping Hsieh, Rabeeh Karimi Mahabadi, Yu-Chun Kao

We study convex optimization problems that feature low-rank matrix solutions. In such scenarios, non-convex methods offer significant advantages over convex methods due to their lower space complexity as well as faster convergence speed. Moreover, many of ...

2018

Finding Mixed Nash Equilibria of Generative Adversarial Networks

Volkan Cevher, Ya-Ping Hsieh, Chen Liu

We reconsider the training objective of Generative Adversarial Networks (GANs) from the mixed Nash Equilibria (NE) perspective. Inspired by the classical prox methods, we develop a novel algorithmic framework for GANs via an infinite-dimensional two-player ...

2018

Mirrored Langevin Dynamics

Volkan Cevher, Paul Thierry Yves Rolland, Ya-Ping Hsieh, Ali Kavis

We consider the problem of sampling from constrained distributions, which has posed significant challenges to both non-asymptotic analysis and algorithmic design. We propose a unified framework, which is inspired by the classical mirror descent, to derive ...

2018

General Proximal Gradient Method: A Case for Non-Euclidean Norms

Volkan Cevher, Ya-Ping Hsieh, Marwa El Halabi, Cong Bang Vu, Van Quang Nguyen

In this paper, we consider composite convex minimization problems. We advocate the merit of considering Generalized Proximal gradient Methods (GPM) where the norm employed is not Euclidean. To that end, we show the tractability of the general proximity ope ...

2017