Illustrative Example Problems

Problem Description

This problem is described in Example 5.1 in [1].

Program Descriptions

  • toyexample.m
    This program runs an interactive demonstration of the convergence of the CE method using Normal updating in 1 dimension. Click the mouse or press a key to advance the display by one iteration.

    Usage:
    Call the program from MATLAB, with the following syntax: toyexample

  • toyexample2.m
    This program runs an alternate interactive demonstration of the convergence of the CE method using Normal updating in 1 dimension. Click the mouse or press a key to advance the display by one iteration.

    Usage:
    Call the program from MATLAB, with the following syntax: toyexample2

  • ssmall.m
    This program illustrates the evolution of the CE method on a very simple problem, showing that it is quite possible to have the standard deviation parameter reduced below an extremely small number.

    Usage:
    Call the program from MATLAB, with the following syntax: sigma = ssmall( N , rho )

    Example: sig = ssmall( 200 , 0.1 )

    Inputs:    
     N - number of samples each iteration
     rho - fraction of best performing samples to take
    Outputs:    
     sigma - A vector of all of the std. deviations

Bibliography

  1. D. P. Kroese and R. Y. Rubinstein.
    The Cross-Entropy Method: A Unified Approach to Combinatorial Optimization, Monte-Carlo Simulation and Machine Learning.
    Springer, 2004.