Hi!
I'm doing some tests with Galapagos to understand better how it works, and I discovered a bit more about how random the solutions are.
It is a small contribution, but maybe you can find it useful:
I don't know if my conclusions are correct or not, but it seems that the solutions are sensible to the start condition.
If the start condition is identical to the solution of the fitness, galapagos, even if randomize the first generations, arrive to the same initial condition. So the result is not random anymore.
I don't know if it is a bug or not the fact that the "randomness" depends on the initial condition... but sure i'm going to study David
post about the lecture of galapagos. Thanks David... it is such a great manual!
Fil
Tags: