Stochastic Frank-Wolfe for Constrained Finite-Sum Minimization

Jul 12, 2020

Speakers

About

We propose a novel Stochastic Frank-Wolfe (≡ Conditional Gradient) algorithm with fixed batch size tailored to the constrained optimization of a finite sum of smooth objectives. We make use of a primal-dual interpretation of the Frank-Wolfe algorithm. Recent work to design stochastic variants of the Frank-Wolfe algorithm fall into two categories: algorithms with increasing batch size, and algorithms with constant batch size. The former have faster convergence rates but are impractical; the latter are practical but slower. Our method combines the advantages of both: it converges for any constant batch size, and has faster theoretical worst case rates than previous fixed batch size algorithms. Our experiments also show faster empirical convergence than previous fixed batch methods for several tasks. Finally, we construct a stochastic estimator of the Frank-Wolfe gap. It allows us to bound the true Frank-Wolfe gap, which is a primal-dual gap in the convex case and a measure of stationarity in general. Our gap estimator can therefore be used as a practical stopping criterion in all cases.

Organizer

Categories

About ICML 2020

The International Conference on Machine Learning (ICML) is the premier gathering of professionals dedicated to the advancement of the branch of artificial intelligence known as machine learning. ICML is globally renowned for presenting and publishing cutting-edge research on all aspects of machine learning used in closely related areas like artificial intelligence, statistics and data science, as well as important application areas such as machine vision, computational biology, speech recognition, and robotics. ICML is one of the fastest growing artificial intelligence conferences in the world. Participants at ICML span a wide range of backgrounds, from academic and industrial researchers, to entrepreneurs and engineers, to graduate students and postdocs.

Store presentation

Should this presentation be stored for 1000 years?

How do we store presentations

Total of 0 viewers voted for saving the presentation to eternal vault which is 0.0%

Sharing

Recommended Videos

Presentations on similar topic, category or speaker