Bi-objective portfolio optimization using Archive Multi-objective Simulated Annealing

Abstract

In the current paper, Bi-objective portfolio optimization problem has been tackled using multiobjective optimization framework. Three popular multiobjective optimization algorithms are used for solving this problem. These are: Archive Multi-objective Simulated Annealing (AMOSA) algorithm, Non-dominated Sorting Genetic algorithm II (NSGA-II) and Multi-objective Particle Swarm Optimization using Crowding distance (MOPSOCD). For each algorithm we trace the Pareto optimal front and compare the results by using four comparisons metrics, Spread, Spacing, Set Coverage and Maximum Spread. Comparative results show that NSGA-II performs the best as compared to the other two algorithms.