On smoothing, regularization, and averaging in stochastic approximation methods for stochastic variational inequality problems

Farzad Yousefian, Angelia Nedich, Uday V. Shanbhag

Research output: Contribution to journalArticle

7 Citations (Scopus)

Abstract

Traditionally, most stochastic approximation (SA) schemes for stochastic variational inequality (SVI) problems have required the underlying mapping to be either strongly monotone or monotone and Lipschitz continuous. In contrast, we consider SVIs with merely monotone and non-Lipschitzian maps. We develop a regularized smoothed SA (RSSA) scheme wherein the stepsize, smoothing, and regularization parameters are reduced after every iteration at a prescribed rate. Under suitable assumptions on the sequences, we show that the algorithm generates iterates that converge to the least norm solution in an almost sure sense, extending the results in Koshal et al. (IEEE Trans Autom Control 58(3):594–609, 2013) to the non-Lipschitzian regime. Additionally, we provide rate estimates that relate iterates to their counterparts derived from a smoothed Tikhonov trajectory associated with a deterministic problem. To derive non-asymptotic rate statements, we develop a variant of the RSSA scheme, denoted by aRSSAr, in which we employ a weighted iterate-averaging, parameterized by a scalar r where r= 1 provides us with the standard averaging scheme. The main contributions are threefold: (i) when r< 1 and the parameter sequences are chosen appropriately, we show that the averaged sequence converges to the least norm solution almost surely and a suitably defined gap function diminishes at an approximate rate O(1/k6) after k steps; (ii) when r< 1 , and smoothing and regularization are suppressed, the gap function admits the rate O(1/k), thus improving the rate O(ln(k)/k) under standard averaging; and (iii) we develop a window-based variant of this scheme that also displays the optimal rate for r< 1. Notably, we prove the superiority of the scheme with r< 1 with its counterpart with r= 1 in terms of the constant factor of the error bound when the size of the averaging window is sufficiently large. We present the performance of the developed schemes on a stochastic Nash–Cournot game with merely monotone and non-Lipschitzian maps.

Original languageEnglish (US)
Pages (from-to)391-431
Number of pages41
JournalMathematical Programming
Volume165
Issue number1
DOIs
StatePublished - Sep 1 2017

Fingerprint

Stochastic Approximation
Variational Inequality Problem
Stochastic Methods
Approximation Methods
Averaging
Smoothing
Regularization
Monotone
Iterate
Gap Function
Approximation Scheme
Trajectories
Converge
Norm
Stochastic Games
Optimal Rates
Smoothing Parameter
Regularization Parameter
Threefolds
Error Bounds

ASJC Scopus subject areas

  • Software
  • Mathematics(all)

Cite this

On smoothing, regularization, and averaging in stochastic approximation methods for stochastic variational inequality problems. / Yousefian, Farzad; Nedich, Angelia; Shanbhag, Uday V.

In: Mathematical Programming, Vol. 165, No. 1, 01.09.2017, p. 391-431.

Research output: Contribution to journalArticle

@article{7a647534db9b4eceb156984ef414d249,
title = "On smoothing, regularization, and averaging in stochastic approximation methods for stochastic variational inequality problems",
abstract = "Traditionally, most stochastic approximation (SA) schemes for stochastic variational inequality (SVI) problems have required the underlying mapping to be either strongly monotone or monotone and Lipschitz continuous. In contrast, we consider SVIs with merely monotone and non-Lipschitzian maps. We develop a regularized smoothed SA (RSSA) scheme wherein the stepsize, smoothing, and regularization parameters are reduced after every iteration at a prescribed rate. Under suitable assumptions on the sequences, we show that the algorithm generates iterates that converge to the least norm solution in an almost sure sense, extending the results in Koshal et al. (IEEE Trans Autom Control 58(3):594–609, 2013) to the non-Lipschitzian regime. Additionally, we provide rate estimates that relate iterates to their counterparts derived from a smoothed Tikhonov trajectory associated with a deterministic problem. To derive non-asymptotic rate statements, we develop a variant of the RSSA scheme, denoted by aRSSAr, in which we employ a weighted iterate-averaging, parameterized by a scalar r where r= 1 provides us with the standard averaging scheme. The main contributions are threefold: (i) when r< 1 and the parameter sequences are chosen appropriately, we show that the averaged sequence converges to the least norm solution almost surely and a suitably defined gap function diminishes at an approximate rate O(1/k6) after k steps; (ii) when r< 1 , and smoothing and regularization are suppressed, the gap function admits the rate O(1/k), thus improving the rate O(ln(k)/k) under standard averaging; and (iii) we develop a window-based variant of this scheme that also displays the optimal rate for r< 1. Notably, we prove the superiority of the scheme with r< 1 with its counterpart with r= 1 in terms of the constant factor of the error bound when the size of the averaging window is sufficiently large. We present the performance of the developed schemes on a stochastic Nash–Cournot game with merely monotone and non-Lipschitzian maps.",
author = "Farzad Yousefian and Angelia Nedich and Shanbhag, {Uday V.}",
year = "2017",
month = "9",
day = "1",
doi = "10.1007/s10107-017-1175-y",
language = "English (US)",
volume = "165",
pages = "391--431",
journal = "Mathematical Programming",
issn = "0025-5610",
publisher = "Springer-Verlag GmbH and Co. KG",
number = "1",

}

TY - JOUR

T1 - On smoothing, regularization, and averaging in stochastic approximation methods for stochastic variational inequality problems

AU - Yousefian, Farzad

AU - Nedich, Angelia

AU - Shanbhag, Uday V.

PY - 2017/9/1

Y1 - 2017/9/1

N2 - Traditionally, most stochastic approximation (SA) schemes for stochastic variational inequality (SVI) problems have required the underlying mapping to be either strongly monotone or monotone and Lipschitz continuous. In contrast, we consider SVIs with merely monotone and non-Lipschitzian maps. We develop a regularized smoothed SA (RSSA) scheme wherein the stepsize, smoothing, and regularization parameters are reduced after every iteration at a prescribed rate. Under suitable assumptions on the sequences, we show that the algorithm generates iterates that converge to the least norm solution in an almost sure sense, extending the results in Koshal et al. (IEEE Trans Autom Control 58(3):594–609, 2013) to the non-Lipschitzian regime. Additionally, we provide rate estimates that relate iterates to their counterparts derived from a smoothed Tikhonov trajectory associated with a deterministic problem. To derive non-asymptotic rate statements, we develop a variant of the RSSA scheme, denoted by aRSSAr, in which we employ a weighted iterate-averaging, parameterized by a scalar r where r= 1 provides us with the standard averaging scheme. The main contributions are threefold: (i) when r< 1 and the parameter sequences are chosen appropriately, we show that the averaged sequence converges to the least norm solution almost surely and a suitably defined gap function diminishes at an approximate rate O(1/k6) after k steps; (ii) when r< 1 , and smoothing and regularization are suppressed, the gap function admits the rate O(1/k), thus improving the rate O(ln(k)/k) under standard averaging; and (iii) we develop a window-based variant of this scheme that also displays the optimal rate for r< 1. Notably, we prove the superiority of the scheme with r< 1 with its counterpart with r= 1 in terms of the constant factor of the error bound when the size of the averaging window is sufficiently large. We present the performance of the developed schemes on a stochastic Nash–Cournot game with merely monotone and non-Lipschitzian maps.

AB - Traditionally, most stochastic approximation (SA) schemes for stochastic variational inequality (SVI) problems have required the underlying mapping to be either strongly monotone or monotone and Lipschitz continuous. In contrast, we consider SVIs with merely monotone and non-Lipschitzian maps. We develop a regularized smoothed SA (RSSA) scheme wherein the stepsize, smoothing, and regularization parameters are reduced after every iteration at a prescribed rate. Under suitable assumptions on the sequences, we show that the algorithm generates iterates that converge to the least norm solution in an almost sure sense, extending the results in Koshal et al. (IEEE Trans Autom Control 58(3):594–609, 2013) to the non-Lipschitzian regime. Additionally, we provide rate estimates that relate iterates to their counterparts derived from a smoothed Tikhonov trajectory associated with a deterministic problem. To derive non-asymptotic rate statements, we develop a variant of the RSSA scheme, denoted by aRSSAr, in which we employ a weighted iterate-averaging, parameterized by a scalar r where r= 1 provides us with the standard averaging scheme. The main contributions are threefold: (i) when r< 1 and the parameter sequences are chosen appropriately, we show that the averaged sequence converges to the least norm solution almost surely and a suitably defined gap function diminishes at an approximate rate O(1/k6) after k steps; (ii) when r< 1 , and smoothing and regularization are suppressed, the gap function admits the rate O(1/k), thus improving the rate O(ln(k)/k) under standard averaging; and (iii) we develop a window-based variant of this scheme that also displays the optimal rate for r< 1. Notably, we prove the superiority of the scheme with r< 1 with its counterpart with r= 1 in terms of the constant factor of the error bound when the size of the averaging window is sufficiently large. We present the performance of the developed schemes on a stochastic Nash–Cournot game with merely monotone and non-Lipschitzian maps.

UR - http://www.scopus.com/inward/record.url?scp=85025831700&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85025831700&partnerID=8YFLogxK

U2 - 10.1007/s10107-017-1175-y

DO - 10.1007/s10107-017-1175-y

M3 - Article

VL - 165

SP - 391

EP - 431

JO - Mathematical Programming

JF - Mathematical Programming

SN - 0025-5610

IS - 1

ER -