0. The genetic approach is a waste of time. It takes too long.
1.Typically, a single hidden layer is sufficient.
2. Minimize the number of hidden nodes by trial and error using a double loop:
a. Outer loop over No. of hidden nodes
b. Inner loop over No. of random initial weight values
3. If you must use genetic, search BOTH the NEWSGROUP and ANSWERS using
greg genetic neural
Hope this helps.