Robin Becker: > interestingly genetic programming et al seem to perform worse when > attempts are made to 'direct' the evolution. Pardon? I thought the point of the cost function was to 'direct' the evolution. Andrew dalke at dalkescientific.com