So, I stumbled across this machine learning algorithm the other day, but can't quite make heads or tails of where it differs from genetic algorithms (in that what can be done in AIS can be done more effectively than with GAs). It seems to me to simply be a GA with fancier names and a more explicit definition of the feedback function. Can anyone elaborate or point me to some papers that can clear up my obviously muddled understanding?

They are very similiar. The difference is the biological inspiration for the algorithms. The actual implementations could be nearly identical. Or they could combine aspects of both, or they could be very different. Even two genetic algorithms can be very different from each other.

Ah, I see. Well, I guess I wasn't too wrong then.

Thanks!

Umm i wouldn't go as far to say they are that similar to a typical genetic algorithm, they certainly have some unique approaches to the 'variation operators'.

Have a look at 'An Immune Algorithm for Protein Structure Prediction on Lattice Models' for an example of one working on a particular problem, also i beleive there is some comparison to a typical GA.

Regards Wolfe