Both theory and a wealth of empirical studies have established that ensembles are more accurate than single predictive models. Unfortunately, the problem of how to maximize ensemble accuracy is, especially for classification, far from solved. This paper presents a novel technique, where genetic algorithms are used for combining several measurements into a complex criterion that is optimized separately for each dataset. The experimental results show that when using the generated combined optimization criteria to rank candidate ensembles, a higher test set accuracy for the top ranked ensemble was achieved compared to using other measures alone, e.g., estimated ensemble accuracy or the diversity measure difficulty.
Sponsorship:
This work was supported by the Information Fusion Research Program (www.infofusion.se) at the University of Skövde, Sweden, in partnership with the Swedish Knowledge Foundation under grant 2003/0104.