A Mean-Variance Optimization algorithm

• DOI: 10.1109/cec.2010.5586027
• Publication Date: 2010-09-29T15:54:29Z
• Authors: Istvan Erlich; Ganesh K. Venayag...; Nakawiro Worawat
• Keywords: Optimization algorithm

A new stochastic optimization algorithm referred to by the authors as `Mean-Variance Optimization' (MVO) is presented in this paper. MVO falls into category of so-called "population-based technique." The uniqueness based on strategic transformation used for mutating offspring mean-variance n-best dynamic population. mapping function transforms uniformly distributed random variation a one characterized variance and mean population attained so far. searching space within restricted range - zero which does not change after applying transformation. Therefore variables are treated always band but evaluation carried out problem range. performance has been demonstrated standard benchmark functions. It shown that finds near optimal solution simple implement. features make it potentially an attractive solving many real-world problems.