Empirical Comparison Of Evolutionary Representations Of The Inverse Problem For Iterated Function Systems

In this paper we present an empirical comparison between evolutionary representations for the resolution of the inverse problem for iterated function systems (IFS). We introduce a class of problem instances that can be used for the comparison of the inverse IFS problem as well as a novel technique that aids exploratory analysis of experiment data. Our comparison suggests that representations that exploit problem specific information, apart from quality/fitness feedback, perform better for the resolution of the inverse problem for IFS.