Adaptive Stochastic Mirror Descent for Constrained Optimization

Anastasia Bayandina

Adaptive Stochastic Mirror Descent for Constrained Optimization

Optimization and Control 2017-05-08 v1

Authors: Anastasia Bayandina

Abstract

Mirror Descent (MD) is a well-known method of solving non-smooth convex optimization problems. This paper analyzes the stochastic variant of MD with adaptive stepsizes. Its convergence on average is shown to be faster than with the fixed stepsizes and optimal in terms of lower bounds.

Keywords

stochastic optimization gradient descent optimization accelerated gradient methods

Cite

@article{arxiv.1705.02031,
  title  = {Adaptive Stochastic Mirror Descent for Constrained Optimization},
  author = {Anastasia Bayandina},
  journal= {arXiv preprint arXiv:1705.02031},
  year   = {2017}
}

Related papers

View all related →