Biomag 96 pp 354-357 | Cite as

Source Localization by Genetic Algorithm

  • R. Takeuchi
  • H. Ikeda
  • A. Ishiyama
  • N. Kasai
Conference paper

Abstract

Recently, it is expected to estimate widely distributed current sources or multiple-dipole sources. Various optimization algorithms are proposed for these estimations, while there must be some difficulties such as local minimization or time consuming problems in those methods. For example, simulated annealing (SA) takes a lot of time to search optimal solutions because this kind of algorithm is not allowed to narrow the search space. Therefore, we consider that genetic algorithm (GA), which is an optimization algorithm based on a mechanics of natural evolution, is effective in rapidly searching the global solution[l][2]. However, specific techniques are necessary to apply GA to biomagnetic source localization. In this paper, we try to improve an application of GA to be more simple, and to assess what effect it will have on these problems.

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. [1]
    Goldberg, D. E. Genetic Algorithms in Search, Optimization, and Machine Learning, Addison-Wesley Publishing Company, 1989.Google Scholar
  2. [2]
    Hatano H. Journal of Japan Artificial Intelligence Conference, 1993, 8–3: 312–319. (in Japanese).Google Scholar

Copyright information

© Springer Science+Business Media New York 2000

Authors and Affiliations

  • R. Takeuchi
    • 1
  • H. Ikeda
    • 1
  • A. Ishiyama
    • 1
  • N. Kasai
    • 2
  1. 1.Department of Electrical EngineeringWaseda UniversityTokyoJapan
  2. 2.Electrotechnical LaboratoryTsukubaJapan

Personalised recommendations