Convergence analysis of gradient descent stochastic algorithms

DOI10.1007/BF02190104MaRDI QIDQ1359455zbMATH OpenOpenAlexFDO

Authors O. Diekmann, Alexander Shapiro, Y. Wardi

Publication date 7 October 1997

Published in Journal of Optimization Theory and Applications (Search for Journal in Brave)

Full work available at URL https://doi.org/10.1007/bf02190104

zbMATH Keywords

discrete event systemsgradient descentinfinitesimaluniform law of large numbers

Mathematics Subject Classification ID

Review text

The paper provides convergence analysis results for a sample-path based stochastic gradient-descent algorithm for optimizing expected-value performance measures in discrete event systems. The algorithm requires that the distance between two consecutive iteration points converge to zero as the iterate count goes to infinity. The paper gives two convergence results: one for the case where the expected-value function is continuously differentiable and the other when the function is nondifferentiable but the sample performance functions are convex. The proofs are based on a version of the uniform law of large numbers which is provable for many discrete event systems where infinitesimal perturbation analysis is known to be strongly consistent.

Recommendations

• Convergence Analysis of Stochastic Algorithms
• Convergence and convergence rate of stochastic gradient search in the case of multiple and non-isolated extrema
• Strong error analysis for stochastic gradient descent optimization algorithms

Cites work

• Asymptotic properties of statistical estimators in stochastic programming
• Convergence Analysis of Stochastic Algorithms
• Convergence of some algorithms for convex minimization
• Convergence of Subdifferentials Under Strong Stochastic Convexity
• scientific article; zbMATH DE number 4013703 (Why is no real title available?)
• scientific article; zbMATH DE number 439380 (Why is no real title available?)
• scientific article; zbMATH DE number 439951 (Why is no real title available?)
• scientific article; zbMATH DE number 50675 (Why is no real title available?)
• scientific article; zbMATH DE number 51708 (Why is no real title available?)
• Interchangeability of expectation and differentiation of waiting times in GI/G/1 queues
• Nondifferentiability of the steady-state function in discrete event dynamic systems
• On sampling controlled stochastic approximation
• Optimization algorithm with probabilistic estimation
• Optimization and nonsmooth analysis
• Optimization of Queues Using an Infinitesimal Perturbation Analysis-Based Stochastic Algorithm with General Update Times
• Optimization via simulation: A review
• Stochastic algorithms with Armijo stepsizes for minimization of functions
• Stochastic Optimization by Simulation: Convergence Proofs for the GI/G/1 Queue in Steady-State
• Stochastic optimization of regenerative systems using infinitesimal perturbation analysis

Cited in

(24)

• Optimization of computer simulation models with rare events
• Convergence of the Monte Carlo expectation maximization for curved exponential families.
• Convergence of stochastic proximal gradient algorithm
• Analysis of biased stochastic gradient descent using sequential semidefinite programs
• Analysis of stochastic gradient descent in continuous time
• Laplacian smoothing gradient descent
• An efficient numerical algorithm for solving data driven feedback control problems
• Accelerated stochastic variance reduction for a class of convex optimization problems
• Retrospective optimization of mixed-integer stochastic systems using dynamic simplex linear interpolation
• Convergence Analysis of Stochastic Algorithms
• ACCELERATING GENERALIZED ITERATIVE SCALING BASED ON STAGGERED AITKEN METHOD FOR ON-LINE CONDITIONAL RANDOM FIELDS
• scientific article; zbMATH DE number 7626722 (Why is no real title available?)
• Plausible Screening Using Functional Properties for Simulations with Large Solution Spaces
• On the convergence analysis of the optimized gradient method
• New Convergence Aspects of Stochastic Gradient Algorithms
• An adaptive gradient method with energy and momentum
• A Sample-Wise Data Driven Control Solver for the Stochastic Optimal Control Problem with Unknown Model Parameters
• SGEM: stochastic gradient with energy and momentum
• PDE-constrained shape optimization: toward product shape spaces and stochastic models
• Spectral projected gradient method for stochastic optimization
• A Global-in-Time Neural Network Approach to Dynamic Portfolio Optimization
• AEGD: adaptive gradient descent with energy
• Assimilating partial observation to enhance feedback control of stochastic dynamical systems
• Performance analysis of stochastic gradient algorithms under weak conditions

This page was built for publication: Convergence analysis of gradient descent stochastic algorithms

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q1359455)