Dynamic fitness inheritance proportion for multi-objective particle swarm optimization

ABSTRACT

In this paper, we propose a dynamic mechanism to vary the probability by which fitness inheritance is applied throughout the run of a multi-objective particle swarm optimizer, in order to obtain a greater reduction in computational cost (than the obtained with a fixed probability), without dramatically affecting the quality of the results. The results obtained show that it is possible to reduce the computational cost by 32% without affecting the quality of the obtained Pareto front.

REFERENCES

• 1: M. Reyes-Sierra and C. A. Coello Coello. Improving PSO-based multi-objective optimization using crowding, mutation and ε-dominance. In Third International Conference on Evolutionary Multi-Criterion Optimization, pages 505--519. LNCS 3410, Springer-Verlag, 2005.
• 2: M. Reyes-Sierra and C. A. Coello Coello. A study of fitness inheritance and approximation techniques for multi-objective particle swarm optimization. In Congress on Evolutionary Computation, pages 65--72. IEEE Service Center, 2005.
• 3: Robert E. Smith , B. A. Dike , S. A. Stegmann, Fitness inheritance in genetic algorithms, Proceedings of the 1995 ACM symposium on Applied computing, p.345-350, February 26-28, 1995, Nashville, Tennessee, United States [doi>10.1145/315891.316014]
• 4: Eckart Zitzler , Kalyanmoy Deb , Lothar Thiele, Comparison of Multiobjective Evolutionary Algorithms: Empirical Results, Evolutionary Computation, v.8 n.2, p.173-195, June 2000 [doi>10.1162/106365600568202]

INDEX TERMS

Primary Classification:
  I. Computing Methodologies
  I.2 ARTIFICIAL INTELLIGENCE
      I.2.8 Problem Solving, Control Methods, and Search
          Subjects: Heuristic methods

Additional Classification:
  G. Mathematics of Computing
  G.1 NUMERICAL ANALYSIS
      G.1.6 Optimization
          Subjects: Global optimization

General Terms:
Algorithms, Performance

Keywords:
fitness inheritance, multi-objective optimization, optimization, particle swarm