English

Fast Algorithms for Online Stochastic Convex Programming

Machine Learning 2014-10-29 v1 Data Structures and Algorithms Optimization and Control

Abstract

We introduce the online stochastic Convex Programming (CP) problem, a very general version of stochastic online problems which allows arbitrary concave objectives and convex feasibility constraints. Many well-studied problems like online stochastic packing and covering, online stochastic matching with concave returns, etc. form a special case of online stochastic CP. We present fast algorithms for these problems, which achieve near-optimal regret guarantees for both the i.i.d. and the random permutation models of stochastic inputs. When applied to the special case online packing, our ideas yield a simpler and faster primal-dual algorithm for this well studied problem, which achieves the optimal competitive ratio. Our techniques make explicit the connection of primal-dual paradigm and online learning to online stochastic CP.

Keywords

Cite

@article{arxiv.1410.7596,
  title  = {Fast Algorithms for Online Stochastic Convex Programming},
  author = {Shipra Agrawal and Nikhil R. Devanur},
  journal= {arXiv preprint arXiv:1410.7596},
  year   = {2014}
}

Comments

To appear in SODA 2015

R2 v1 2026-06-22T06:38:33.814Z