Hier finden Sie wissenschaftliche Publikationen aus den Fraunhofer-Instituten.

A maximum entropy approach to sampling in EDA - the single connected case

: Ochoa, A.; Höns, R.; Soto, M.; Mühlenbein, H.

Sanfeliu, A.:
Progress in pattern recognition, speech and image analysis. 8th Iberoamerican Congress on Pattern Recognition : Havana, Cuba, November 26-29, 2003. Proceedings, CIARP 2003
Berlin: Springer, 2003 (Lecture Notes in Computer Science 2905)
ISBN: 3-540-20590-X
ISSN: 0302-9743
Iberoamerican Congress on Pattern Recognition (CIARP) <8, 2003, Havana>
Conference Paper
Fraunhofer AIS ( IAIS) ()

The success of evolutionary algorithms, in particular Factorized Distribution Algorithms (FDA), for many pattern recognition tasks heavily depends on our ability to reduce the number of function evaluations. This paper introduces a method to reduce the population size overhead. We use low order maxginals during the learning step and then compute the maximum entropy joint distributions for the cliques of the graph. The maximum entropy distribution is computed by an Iterative Proportional Fitting embedded in a junction tree message passing scheme to ensure consistency. We show for the class of single connected FDA that our method outperforms the commonly-used PLS sampling.