MOCEO: A Proposal for Multiple Objective Cross-Entropy Optimization Method

We provide a novel Cross-Entropy optimization approach solving multi-objective optimization problems, that is called Multi-Objective Cross-Entropy Optimization (MOCEO) in recent article. The Cross-Entropy (CE) method belongs to one kind of the stochastic learning algorithm, which is inspired from the rare event simulation problems, and is proved to be successful and converge quickly in the case of single objective otimization problems. Our study modifies the basic CE method and extends the application of the algorithm for solving multi-objective optimization problems. A new parameter updating mechanism is used in MOCEO, and a recombination operator is implemented in MOCEO to enhance the algorithm´s global search ability. In order to maintain the diversity of the population and to improve the computational efficiency, two truncation mechanisms for individual selection are applied in the algorithm. MOCEO has been evaluated on some standard multi-objective optimization test problems and the performance assessed by using different performance metrics. Comparing to some well-known multi-objective evolutionary algorithms and with recently proposed multi-objective Cross-Entropy algorithms, the simulation results demonstrate that the MOCEO is an effective algorithm for solving multi-object optimization problems.