Application to Artificial Intelligence

Genetic algorithms have been applied to almost any kind of problem, such as, object recognition for artificial vision [SCC97,BLVY02], robotics path planing [Dav91,SAM⁺04], parameter estimation [JH90,OBHMMP03], instance selection [CHL03,Kun95], reinforcement learning [MSG99], and neural network [MTH91,AT93,BGK97] and ensemble design [ZWT02].

Real-coded genetic algorithms using CIXL2 can be applied to any of these problems provided they are defined in a continuous domain. We have chosen an application of RCGAs to the estimation of the weight of each network in an ensemble. This is an interesting problem where standard methods encounter many difficulties.