Research of the dissimilation strategy for MEBML

MEBML, mind-evolution-based machine learning mainly consists of similar taxis and dissimilation operators. Especially, the dissimilation strategy has important effects on the evolution efficiency and global optimality. In the paper, five dissimilation strategies based on the analysis of effects of the dissimilation operator in MEBML and the dissimilation mechanism. Finally, comparisons of these dissimilation strategies are made through the example of a global optimization problem.