Performance Improvement of Network Based Parallel Genetic Algorithm by Exploiting Server's Computing Power

서버의 계산능력을 활용한 네트워크기반 병렬유전자알고리즘의 성능향상

  • 송봉기 (부경대학교 정보시스템) ;
  • 김용성 (경남정보대학 컴퓨터정보계) ;
  • 성길영 (경상대학교 정보통신공학) ;
  • 우종호 (부경대학교 전자컴퓨터정보통신공학부)
  • Published : 2004.07.01

Abstract

This paper proposes a method improving the convergence speed of optimal solution for parallel genetic algorithm in the network based client-server model. Unlike the existing methods of obtaining global elite only by evaluating local elites in server, the proposed method obtains it by evaluating local elites and improving its fitness by applying genetic algorithm during idle time of the server. By using the improved chromosome in server for the client's genetic algorithm processing, the convergence speed of the optimal solution is increased. The improvement of fitness at the server during the interval of chromosome migration is (equation omitted)(F$_{max}$(g)-F$_{max}$(g-1)), whole F$_{max}$(g) is a max fitness of the g-th generation and G is the number of improved generation by the server. As the number of clients increases and G decreases, the improvement of fitness goes down. However the improvement of fitness is better than existing methods..

본 논문에서는 네트워크기반의 클라이언트-서버모델에서 병렬유전자알고리즘의 최적해 수렴속도를 향상시키는 방법을 제안한다. 전역 최적해를 지역 엘리트의 평가만으로 구하는 기존의 방법과는 달리 제안한 방법은 서버에서 지역 엘리트의 평가를 통해 전역 최적해를 구하고 유휴시간에 유전자알고리즘을 적용하여 전역 최적해의 적합도를 개선한다. 서버에서 개선된 전역 최적해를 클라이언트의 유전자알고리즘에서 사용하므로 전체 알고리즘의 최적해 수렴속도가 향상된다 Fmax(g)는 g번째 세대의 최대 적합도, G는 서버에서 개선되는 세대수일 때, 지역 최적해의 이주 시 서버에서 개선되는 적합도는 (equation omitted)(F/sub max/(g)-F/sub max/(g-1)) 이다. 여기서 클라이언트의 수가 증가하면 G가 작아져서 적합도 개선치는 줄어드나 기존의 방법보다 적합도가 개선됨을 확인할 수 있었다.

Keywords

References

  1. J.H. Holland, 'Adaptation in Natural and Artificial Systems', University of Michigan Press, 1975
  2. H. Kitano, 'Empirical Studies on the Speed of Convergence of the Neural Network Training by Genetic Algorithm', Proc. of AAAI-90, 1990
  3. L. Wang, A.A. Maciejewski, H.J. Siegel and V.P. Roychowdhury, 'A Comparative Study of Five Parallel Genetic Algorithms Using the Traveling Salesman Problem', 1998 IPPS/SPDP. Proc. of the First Merged International. 1998. p.345-349, 1998. 4. https://doi.org/10.1109/IPPS.1998.669938
  4. M. Nowostawski and R. Poli, 'Parallel Genetic Algorithm Taxonomy', International Conference on Knowledge-Based Intelligent Information Engineering Systems, p.88-92, 1999. 8 https://doi.org/10.1109/KES.1999.820127
  5. M. Golub and D. Jakobovic, 'A New Model Of Global Parallel Genetic Algorithm', Information Technology Interfaces ITI 2000, p.363-368, 2000. 1
  6. E. Cantu-Paz, 'Designing Efficient Master-Slave Parallel Genetic Algorithms', Genetic Programming: Proc. of the Third Annual conference, p.455-460, 1998
  7. L. Tan and K.A. Smith, 'A New Parallel Genetic Algorithm', Proc. of the International Symposium on Parallel Architectures, Algorithm and Networks, IEEE, 2002 https://doi.org/10.1109/ISPAN.2002.1004301
  8. K. Kojima, W. Kawamata, H. Matsuo and M. Ishigame, 'Network based Parallel Genetic Algorithm using Client-Server Model', Proc. of CEC2002, p.244-249, 2000 https://doi.org/10.1109/CEC.2000.870302
  9. K. Kojima, H. Matsuo and M. Ishigame, 'Reduction of Communication Quantity for Network Based Parallel GA', Proc. of the CEC2002 Congress on , Volume: 2, p.1715-1720, 2002. 5 https://doi.org/10.1109/CEC.2002.1004501
  10. D. E. Goldberg, Genetic Algorithms in Search, Optimization and Machine Learning, Addison-Wesley, New York, NY, 1989
  11. J. Cheng, W. Chen, L, Chen and Y. Ma, 'The Improvement of Genetic Algorithm Searching Performance', Proc. of the First International Conference on Machine Learning and Cybernetics, Beijing, p.947-951, 2002. 11 https://doi.org/10.1109/ICMLC.2002.1174523