A Multiobjective Memetic Algorithm for Multiobjective Unconstrained Binary Quadratic Programming Problem

This study introduces a multiobjective memetic algorithm for multiobjective unconstrained binary quadratic programming problem (mUBQP). It integrates multiobjective evolutionary algorithm based on decomposition and tabu search to search an approximate Pareto front with good convergence and diversity. To further enhance the search ability, uniform generation is introduced to generate different uniform weight vectors for decomposition in every generation. The proposed algorithm is tested on 50 mUBQP instances. Experimental results show the effectiveness of the proposed algorithm in solving mUBQP.