A New Virtual Population Based Incremental Learning Approach for Optimizations Using Selfish Gene Theory

Feng Wang,Yuanxiang Li
DOI: https://doi.org/10.1109/ICNC.2009.577
2009-01-01
Abstract:In this paper, we proposed a new approach which employed the selfish gene theory to construct virtual population for optimizations. And an incremental learning scheme which based on mutual information entropy was also used to speed up the convergence velocity. Experimental results on several benchmark problems show that, this new approach often performs better than BMDA, COMIT and MIMIC.
What problem does this paper attempt to address?