The case for and against fixed step-size: Stochastic approximation algorithms in optimization and machine learning
Fuente:
arXiv
Guardado en:
| Autores principales: | , , |
|---|---|
| Formato: | Preprint |
| Publicado: |
2023
|
| Materias: | |
| Acceso en línea: | |
| Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| _version_ | 1866917071740207104 |
|---|---|
| author | Lauand, Caio Kalil Kontoyiannis, Ioannis Meyn, Sean |
| author_facet | Lauand, Caio Kalil Kontoyiannis, Ioannis Meyn, Sean |
| contents | Theory and application of stochastic approximation (SA) have become increasingly relevant due in part to applications in optimization and reinforcement learning. This paper takes a new look at SA with constant step-size $α>0$, defined by the recursion, $$θ_{n+1} = θ_{n}+ αf(θ_n,Φ_{n+1})$$ in which $θ_n\in\mathbb{R}^d$ and $\{Φ_{n}\}$ is a Markov chain. The goal is to approximately solve root finding problem $\bar{f}(θ^*)=0$, where $\bar{f}(θ)=\mathbb{E}[f(θ,Φ)]$ and $Φ$ has the steady-state distribution of $\{Φ_{n}\}$.
The following conclusions are obtained under an ergodicity assumption on the Markov chain, compatible assumptions on $f$, and for $α>0$ sufficiently small:
$\textbf{1.}$ The pair process $\{(θ_n,Φ_n)\}$ is geometrically ergodic in a topological sense.
$\textbf{2.}$ For every $1\le p\le 4$, there is a constant $b_p$ such that $\limsup_{n\to\infty}\mathbb{E}[\|θ_n-θ^*\|^p]\le b_p α^{p/2}$ for each initial condition.
$\textbf{3.}$ The Polyak-Ruppert-style averaged estimates $θ^{\text{PR}}_n=n^{-1}\sum_{k=1}^{n}θ_k$ converge to a limit $θ^{\text{PR}}_\infty$ almost surely and in mean square, which satisfies $θ^{\text{PR}}_\infty=θ^*+α\barΥ^*+O(α^2)$ for an identified non-random $\barΥ^*\in\mathbb{R}^d$. Moreover, the covariance is approximately optimal: The limiting covariance matrix of $θ^{\text {PR}}_n$ is approximately minimal in a matricial sense.
The two main take-aways for practitioners are application-dependent. It is argued that, in applications to optimization, constant gain algorithms may be preferable even when the objective has multiple local minima; while a vanishing gain algorithm is preferable in applications to reinforcement learning due to the presence of bias. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2309_02944 |
| institution | arXiv |
| publishDate | 2023 |
| record_format | arxiv |
| spellingShingle | The case for and against fixed step-size: Stochastic approximation algorithms in optimization and machine learning Lauand, Caio Kalil Kontoyiannis, Ioannis Meyn, Sean Statistics Theory Machine Learning 62L20, 68T05 Theory and application of stochastic approximation (SA) have become increasingly relevant due in part to applications in optimization and reinforcement learning. This paper takes a new look at SA with constant step-size $α>0$, defined by the recursion, $$θ_{n+1} = θ_{n}+ αf(θ_n,Φ_{n+1})$$ in which $θ_n\in\mathbb{R}^d$ and $\{Φ_{n}\}$ is a Markov chain. The goal is to approximately solve root finding problem $\bar{f}(θ^*)=0$, where $\bar{f}(θ)=\mathbb{E}[f(θ,Φ)]$ and $Φ$ has the steady-state distribution of $\{Φ_{n}\}$. The following conclusions are obtained under an ergodicity assumption on the Markov chain, compatible assumptions on $f$, and for $α>0$ sufficiently small: $\textbf{1.}$ The pair process $\{(θ_n,Φ_n)\}$ is geometrically ergodic in a topological sense. $\textbf{2.}$ For every $1\le p\le 4$, there is a constant $b_p$ such that $\limsup_{n\to\infty}\mathbb{E}[\|θ_n-θ^*\|^p]\le b_p α^{p/2}$ for each initial condition. $\textbf{3.}$ The Polyak-Ruppert-style averaged estimates $θ^{\text{PR}}_n=n^{-1}\sum_{k=1}^{n}θ_k$ converge to a limit $θ^{\text{PR}}_\infty$ almost surely and in mean square, which satisfies $θ^{\text{PR}}_\infty=θ^*+α\barΥ^*+O(α^2)$ for an identified non-random $\barΥ^*\in\mathbb{R}^d$. Moreover, the covariance is approximately optimal: The limiting covariance matrix of $θ^{\text {PR}}_n$ is approximately minimal in a matricial sense. The two main take-aways for practitioners are application-dependent. It is argued that, in applications to optimization, constant gain algorithms may be preferable even when the objective has multiple local minima; while a vanishing gain algorithm is preferable in applications to reinforcement learning due to the presence of bias. |
| title | The case for and against fixed step-size: Stochastic approximation algorithms in optimization and machine learning |
| topic | Statistics Theory Machine Learning 62L20, 68T05 |
| url | https://arxiv.org/abs/2309.02944 |