Towards Gradient Free and Projection Free Stochastic Optimization
Anit Kumar Sahu, Manzil Zaheer, Soummya Kar
Abstract
This paper focuses on the problem of \emph{constrained} \emph{stochastic} optimization. A zeroth order Frank-Wolfe algorithm is proposed, which in addition to the projection-free nature of the vanilla Frank-Wolfe algorithm makes it gradient free. Under convexity and smoothness assumption, we show that the proposed algorithm converges to the optimal objective function at a rate $O\left(1/T^{1/3}\right)$, where $T$ denotes the iteration count. In particular, the primal sub-optimality gap is shown to have a dimension dependence of $O\left(d^{1/3}\right)$, which is the best known dimension dependence among all zeroth order optimization algorithms with one directional derivative per iteration. For non-convex functions, we obtain the \emph{Frank-Wolfe} gap to be $O\left(d^{1/3}T^{-1/4}\right)$. Experiments on black-box optimization setups demonstrate the efficacy of the proposed algorithm.
BibTeX
@InProceedings{pmlr-v89-sahu19a,
title = {Towards Gradient Free and Projection Free Stochastic Optimization},
author = {Sahu, Anit Kumar and Zaheer, Manzil and Kar, Soummya},
booktitle = {Proceedings of the Twenty-Second International Conference on Artificial Intelligence and Statistics},
pages = {3468--3477},
year = {2019},
editor = {Chaudhuri, Kamalika and Sugiyama, Masashi},
volume = {89},
series = {Proceedings of Machine Learning Research},
month = {16--18 Apr},
publisher = {PMLR},
pdf = {http://proceedings.mlr.press/v89/sahu19a/sahu19a.pdf},
url = {https://proceedings.mlr.press/v89/sahu19a.html},
abstract = {This paper focuses on the problem of \emph{constrained} \emph{stochastic} optimization. A zeroth order Frank-Wolfe algorithm is proposed, which in addition to the projection-free nature of the vanilla Frank-Wolfe algorithm makes it gradient free. Under convexity and smoothness assumption, we show that the proposed algorithm converges to the optimal objective function at a rate $O\left(1/T^{1/3}\right)$, where $T$ denotes the iteration count. In particular, the primal sub-optimality gap is shown to have a dimension dependence of $O\left(d^{1/3}\right)$, which is the best known dimension dependence among all zeroth order optimization algorithms with one directional derivative per iteration. For non-convex functions, we obtain the \emph{Frank-Wolfe} gap to be $O\left(d^{1/3}T^{-1/4}\right)$. Experiments on black-box optimization setups demonstrate the efficacy of the proposed algorithm.}
}