Stochastic Search and Optimization

Code in support of book Introduction to Stochastic Search and Optimization.

Description

Introduction to Stochastic Search and Optimization is an overview of the principles, algorithms, and practical aspects of stochastic optimization, including applications drawn from engineering, statistics, and computer science. The book may serve as either a reference book for researchers and practitioners or as a textbook, the latter use being supported by exercises at the end of every chapter and appendix. The text covers a broad range of the most widely used stochastic methods, including:

Random search· Recursive linear estimation· Stochastic approximation· Simulated annealing· Genetic and evolutionary algorithms· Machine (reinforcement) learning· Model selection· Simulation-based optimization· Markov chain Monte Carlo· Optimal experimental design
The MATLAB code here is in support of the book. Additional information on the book and MATLAB code is available at http://www.jhuapl.edu/ISSO/

• MATLAB release: MATLAB 6.1 (R12.1)

Other Files

twospsaconstrained.m,
basicRM_SA.m,
bit2num.m,
FDSA.m,
fitpop.m,
GAbit_roulette.m,
GAbit_tourney.m,
gainsFDSA.m,
gainsSPSA.m,
GAmean_distance.m,
GAreal_tourney.m,
globspsa.m,
iterate_avg.m,
num2bit.m,
randsrchA.m,
randsrchABC.m,
randsrchB_noise.m,
simannel.m,
spsa_basic_constrained.m,
spsa_withoptions.m,
SPvsFD.m,
twoSGconstrained.m

[chandran Ponnuswamy]

 

[stacy hill]

Given its use with book, I assumed the code was not intended to be production code nor necessarily the most efficient, but rather was designed to meet the pedagogical aims of the book (I also own a copy of the book). The availability of these algorithms provides a valuable learning tool and supplement to the book. Given this objective, the code demonstrates the algorithms covered in the book and appear perfectly suitable for moderate-size problems.

[salah eddine]

 

[Erman Ozguven]