Computer Science Home Page
However, in the last few years the usefulness of schemata has been widely criticised, and the attention of GA theorists has moved away from schemata to land onto Markov chains. To be very clear: quite a number of researchers nowadays believe that schema theorems are nothing more than trivial tautologies of no use whatsoever.
In this talk, after explaining what schemata and schema theorems are, I will show why this is incorrect and why one should be interested in schema theories. I will do that by presenting some very recent results including Chris Stephens exact schema theory for GAs, my extension of it to genetic programming, the concept of effective fitness in GAs and GP, and, time permitting, some other schema-based results on the convergence and population sizing of GAs.