Has anybody got any information about algorithms that can 'learn' about the environment that they control. In other words optimise the variables that control the algorithm for a given function. Any leads on books, web sites, etc would be appreciated. Tony Just when I thought I knew it all, I learned that I didn't.