Stochastic recursive algorithms for optimization. Simultaneous perturbation methods
algorithms of Robbins-Monro and Kiefer-Wolfowitzgradient schemesMarkov decision processstochastic recursive algorithms
Research exposition (monographs, survey articles) pertaining to operations research and mathematical programming (90-02) Traffic problems in operations research (90B20) Stochastic scheduling theory in operations research (90B36) Quadratic programming (90C20) Markov and semi-Markov decision processes (90C40)
A general optimization problem considered in the present book has the form: Find \(\theta\in\mathbb{R}^n\) that minimizes \(J(\theta)\), \(\theta\in C\). If one has complete information about \(J\) and \(C\), this is a well-known deterministic optimization problem. If \(J\) is the expected value over noisy observations, then it is a stochastic optimization problem. Stochastic approximations have its roots in a famous paper by \textit{H. Robbins} and \textit{S. Monro} [Ann. Math. Stat. 22, 400--407 (1951; Zbl 0054.05901)]. \textit{J. Kiefer} and \textit{J. Wolfowitz} [Ann. Math. Stat. 23, 462--466 (1952; Zbl 0049.36601)] published the first stochastic approximation algorithm for stochastic optimization. The book under review summarizes the recent research on simultaneously perturbation problems. Many results of the text mainly based on the author's own contributions to this topic. The book provides a coverage of the known material in stochastic optimizations, such that both researchers and practitioners should find it useful. The text is well understandable, the book is clearly written and impressively printed. Theorems and algorithms are emphasized in coloured frames. Therefore, the book can be used as material for lectures dedicated to master students. The book is divided into 6 parts (14 chapters): 1. Introduction to stochastic recursive algorithms, -- 2. Gradient estimation schemes, -- 3. Hessian estimation schemes, -- 4. Variations to the basic scheme, -- 5. Applications, -- 6. Appendix. There are references at the end of every chapter. The appendix informs on martingales, ODE, the Borkar-Meyn theorem and the Kushner-Clark theorem. Applications are studied in service systems, road traffic control and in communication networks.
- scientific article; zbMATH DE number 53271
- scientific article; zbMATH DE number 1043533
- scientific article; zbMATH DE number 194340
- Biologically inspired optimization methods: An introduction.
- Smoothed Functionals and Constrained Stochastic Approximation
- Stochastic approximation algorithm for minimax problems
- Stochastic multi-stage optimization. At the crossroads between discrete time stochastic control and stochastic programming
- Decreasing step Stochastic algorithms: a.s. behaviour of weighted empirical measures
- Asymptotic optimality in stochastic optimization
- Variance-constrained actor-critic algorithms for discounted and average reward MDPs
- Nested kriging predictions for datasets with a large number of observations
- Generalization of a result of Fabian on the asymptotic normality of stochastic approximation
- Stochastic optimization of forward recursive functions
- Recurrent neural networks as optimal mesh refinement strategies
- Simulation methods for robust risk assessment and the distorted mix approach
- Quantum simulation of the ground-state Stark effect in small molecules: a case study using IBM Q
- Smoothed functional-based gradient algorithms for off-policy reinforcement learning: a non-asymptotic viewpoint
- Simultaneous perturbation Newton algorithms for simulation optimization
- On Choosing Parameters in Retrospective-Approximation Algorithms for Stochastic Root Finding and Simulation Optimization
- Multiscale Q-learning with linear function approximation
- Stochastic optimisation with inequality constraints using simultaneous perturbations and penalty functions
- Recursive Stochastic Algorithms for Global Optimization in $\mathbb{R}^d $
- Introduction to Stochastic Search and Optimization
- Optimal random perturbations for stochastic approximation using a simultaneous perturbation gradient approximation
- scientific article; zbMATH DE number 1784873 (Why is no real title available?)
- Risk-constrained reinforcement learning with percentile risk criteria
- Newton-based stochastic optimization using \(q\)-Gaussian smoothed functional algorithms
- Iterative learning control using faded measurements without system information: a gradient estimation approach
- Risk-Sensitive Reinforcement Learning via Policy Gradient Search
- Gradient-based adaptive stochastic search for simulation optimization over continuous space
- A simulation-based algorithm for optimal pricing policy under demand uncertainty
- Derivative-free optimization methods
- Quasi-Newton smoothed functional algorithms for unconstrained and constrained simulation optimization
- Actor-critic algorithms with online feature adaptation
- Variable ansatz applied to spectral operator decomposition in a physical superconducting quantum device
- A model for data transmission and its optimization
- Stochastic approximation procedures for Lévy-driven SDEs
- Modeling and control of data transmission
- Learning equilibrium mean‐variance strategy
- Truncated Cauchy random perturbations for smoothed functional-based stochastic optimization
- Optimal experimental design: formulations and computations
- Risk-averse constrained blackbox optimization under mixed aleatory/epistemic uncertainties
- Active nodes maximization in a virus spread model: an SI2R malware propagation model
- A pseudo-gradient approach for model-free Markov chain optimization
This page was built for publication: Stochastic recursive algorithms for optimization. Simultaneous perturbation methods
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q441138)
