• Home
  • Search
  • Stochastic Approximation Beyond Gradient for Signal Processing and Machine Learning
  • Cite Icon10
  • https://doi.org/10.1109/tsp.2023.3301121Copy DOI Icon

Stochastic Approximation Beyond Gradient for Signal Processing and Machine Learning

Show More
  • Abstract
  • Literature Map
  • References
  • Citations
  • Similar Papers
Abstract

Stochastic Approximation ( <bold xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">SA</b> ) is a classical algorithm that has had since the early days a huge impact on signal processing, and nowadays on machine learning, due to the necessity to deal with a large amount of data observed with uncertainties. An exemplar special case of <bold xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">SA</b> pertains to the popular stochastic (sub)gradient algorithm which is the working horse behind many important applications. A lesser-known fact is that the <bold xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">SA</b> scheme also extends to non-stochastic-gradient algorithms such as compressed stochastic gradient, stochastic expectation-maximization, and a number of reinforcement learning algorithms. The aim of this article is to overview and introduce the non-stochastic-gradient perspectives of <bold xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">SA</b> to the signal processing and machine learning audiences through presenting a design guideline of <bold xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">SA</b> algorithms backed by theories. Our central theme is to propose a general framework that unifies existing theories of <bold xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">SA</b> , including its non-asymptotic and asymptotic convergence results, and demonstrate their applications on popular non-stochastic-gradient algorithms. We build our analysis framework based on classes of Lyapunov functions that satisfy a variety of mild conditions. We draw connections between non-stochastic-gradient algorithms and scenarios when the Lyapunov function is smooth, convex, or strongly convex. Using the said framework, we illustrate the convergence properties of the non-stochastic-gradient algorithms using concrete examples. Extensions to the emerging variance reduction techniques for improved sample complexity will also be discussed.

Similar Papers
  • Research Article
  • Citations31

Biased stochastic conjugate gradient algorithm with adaptive step size for nonconvex problems

  • Sep 22, 2023
  • Expert Systems with Applications
  • Ruping Huang +3
  • Conference Article
  • Citations7

A Novel Constrained Adaptive Algorithm Using the Conjugate Gradient Method for Smart Antennas

  • Nov 01, 2007
  • Lei Wang +1
  • Research Article
  • Citations3

Adaptive Biased Stochastic Optimization.

  • Jan 01, 2025
  • IEEE transactions on pattern analysis and machine intelligence
  • Zhuang Yang
  • Research Article

Large-Scale Phase Retrieval via Stochastic Reweighted Amplitude Flow

  • Nov 30, 2020
  • KSII Transactions on Internet and Information Systems
  • Zhuolei Xiao +2
  • Research Article
  • Citations19

Adaptive filtering‐based multi‐innovation gradient algorithm for input nonlinear systems with autoregressive noise

  • Apr 17, 2017
  • International Journal of Adaptive Control and Signal Processing
  • Yawen Mao +2
  • Conference Article

Low complexity variable step size mechanisms for constrained minimum variance CDMA receivers

  • May 17, 2004
  • R.C De Lamare +1
  • Conference Article
  • Citations14

Scalable Semi-Supervised SVM via Triply Stochastic Gradients

  • Aug 01, 2019
  • Xiang Geng +5
  • Research Article
  • Citations46

Averaged stochastic gradient algorithms for adaptive blind multiuser detection in DS/CDMA systems

  • Jan 01, 2000
  • IEEE Transactions on Communications
  • V Krishnamurthy
  • Conference Article

A Kalman smoother based estimation approach for dynamic adjustment systems with irregularly missing output data

  • May 10, 2021
  • Shaohua Ding +3
  • Conference Article

Several Stochastic Gradient Algorithms for Nonlinear Systems with Hard Nonlinearities

  • Aug 01, 2015
  • Jia Tang
  • Research Article
  • Citations12

Utility-Optimal Wireless Routing in the Presence of Heavy Tails

  • Jan 01, 2019
  • IEEE Transactions on Vehicular Technology
  • Shuang Xia +2
  • Research Article
  • Citations4

Multikernel Passive Stochastic Gradient Algorithms and Transfer Learning

  • May 11, 2021
  • IEEE Transactions on Automatic Control
  • Vikram Krishnamurthy +1
  • Research Article
  • Citations3

A comparison of stochastic gradient and minimum entropy deconvolution algorithms

  • Sep 01, 1988
  • Signal Processing
  • A.T Walden
  • Conference Article
  • Citations374

Space or time adaptive signal processing by neural network models

  • Jan 01, 1986
  • AIP conference proceedings
  • J Herault +1
  • Research Article
  • Citations141

On the design of gradient algorithms for digitally implemented adaptive filters

  • Jan 01, 1973
  • IEEE Transactions on Circuit Theory
  • R Gitlin +2
Cactus Communications logo

Copyright 2026 Cactus Communications. All rights reserved.