Comparison Of Methods For Using Reduced Models To Speed Up Design Optimization

In this paper we compare two methods for forming reduced models to speed up geneticalgorithm-based optimization. The methods work by forming functional approximations of the tness function which are used to speed up the GA optimization. One method speeds up the optimization by making the genetic operators more informed. The other method speeds up the optimization by genetically engineering some individuals instead of using the regular Darwinian evolution approach. Empirical results in several engineering design domains are presented.