Diversity metric assessment for multi-objective optimization

In multi-objective optimization evolutionary algorithms (MOEA) diversity is an important issue and many researchers have investigated this notion. Due to the importance of multi-objective optimization in industry and engineering, it is essential to find a diverse set of Pareto optimal solutions which covers as much space as possible in the feasible region of the solution space. Theresearch (2012) presented in this paper aims to examine the performance and compare two metrics for diversity assessment of the non-dominated solution (i.e. Pareto set) in the solution space. These two metrics have been programmed in (MATLAB) software and implemented in number of simple bi-objective optimization cases to compare the analytical results with the visual distribution of points. The experimental results show that the circumscription diversity (CM) metric outperforms the Pair Wise (PW) metric