Extremely Fast Convergence Rates for Extremum Seeking Control with Polyak-Ruppert Averaging

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Lauand, Caio Kalil, Meyn, Sean
Natura: Preprint
Pubblicazione: 2022
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866911810964160512
author Lauand, Caio Kalil
Meyn, Sean
author_facet Lauand, Caio Kalil
Meyn, Sean
contents Stochastic approximation is a foundation for many algorithms found in machine learning and optimization. It is in general slow to converge: the mean square error vanishes as $O(n^{-1})$. A deterministic counterpart known as quasi-stochastic approximation is a viable alternative in many applications, including gradient-free optimization and reinforcement learning. It was assumed in prior research that the optimal achievable convergence rate is $O(n^{-2})$. It is shown in this paper that through design it is possible to obtain far faster convergence, of order $O(n^{-4+δ})$, with $δ>0$ arbitrary. Two techniques are introduced for the first time to achieve this rate of convergence. The theory is also specialized within the context of gradient-free optimization, and tested on standard benchmarks. The main results are based on a combination of novel application of results from number theory and techniques adapted from stochastic approximation theory.
format Preprint
id arxiv_https___arxiv_org_abs_2206_00814
institution arXiv
publishDate 2022
record_format arxiv
spellingShingle Extremely Fast Convergence Rates for Extremum Seeking Control with Polyak-Ruppert Averaging
Lauand, Caio Kalil
Meyn, Sean
Optimization and Control
62L20, 34C29
Stochastic approximation is a foundation for many algorithms found in machine learning and optimization. It is in general slow to converge: the mean square error vanishes as $O(n^{-1})$. A deterministic counterpart known as quasi-stochastic approximation is a viable alternative in many applications, including gradient-free optimization and reinforcement learning. It was assumed in prior research that the optimal achievable convergence rate is $O(n^{-2})$. It is shown in this paper that through design it is possible to obtain far faster convergence, of order $O(n^{-4+δ})$, with $δ>0$ arbitrary. Two techniques are introduced for the first time to achieve this rate of convergence. The theory is also specialized within the context of gradient-free optimization, and tested on standard benchmarks. The main results are based on a combination of novel application of results from number theory and techniques adapted from stochastic approximation theory.
title Extremely Fast Convergence Rates for Extremum Seeking Control with Polyak-Ruppert Averaging
topic Optimization and Control
62L20, 34C29
url https://arxiv.org/abs/2206.00814