The Transactions of the Korean Institute of Electrical Engineers D (대한전기학회논문지:시스템및제어부문D)

The Korean Institute of Electrical Engineers (대한전기학회)
권호 표지

Using Higher Order Neuron on the Supervised Learning Machine of Kohonen Feature Map

고차 뉴런을 이용한 교사 학습기의 Kohonen Feature Map

  • Published : 2003.05.01
Download PDF
⟨ Previous Next ⟩

Abstract

In this paper we propose Using Higher Order Neuron on the Supervised Learning Machine of the Kohonen Feature Map. The architecture of proposed model adopts the higher order neuron in the input layer of Kohonen Feature Map as a Supervised Learning Machine. It is able to estimate boundary on input pattern space because or the higher order neuron. However, it suffers from a problem that the number of neuron weight increases because of the higher order neuron in the input layer. In this time, we solved this problem by placing the second order neuron among the higher order neuron. The feature of the higher order neuron can be mapped similar inputs on the Kohonen Feature Map. It also is the network with topological mapping. We have simulated the proposed model in respect of the recognition rate by XOR problem, discrimination of 20 alphabet patterns, Mirror Symmetry problem, and numerical letters Pattern Problem.

Keywords

References

  1. Remelhart and Mcclelland, 'Parallel Distributed Processing', MIT, Press, 1986
  2. T. Kohonen, 'Self-Origanized formation of topographicall correct feature map', Biol, Cybern, pp.59-69, 43, 1982 https://doi.org/10.1007/BF00337288
  3. T. Kohonen, 'Self-Origanization and Associative Memory: Second edition', Springer-Verlag, Press, 1988
  4. T. Kohonen, 'Self-Origanizing Semantic Maps', Biol, Cyber, 61, pp.241-253, 198 https://doi.org/10.1007/BF00203171
  5. H. Ichiki, M.Hagiwara and M. Nakagawa, 'Self-Organizing Multi-Layer Semantic Maps', IJCNN, Vol.1, pp.357-360, 1991
  6. H. Ichiki, M.Hagiwara and M. Nakagawa, 'Kohonen Feature Map as a Supervised Learning Machine', ICNN, pp.1944-1948, 1993 https://doi.org/10.1109/ICNN.1993.298854
  7. Y. C. Shin and R. Sridhar, 'Network Congigurations and Training Speeds of Second-Order Neural Network', WCONN, Vol.1, pp.585-588, 1993
  8. J. G. Taylor and S. Coonnber, 'Learning Higher-Order correlation', Neural Network, Vol.6, pp.423-427, 1993 https://doi.org/10.1016/0893-6080(93)90009-L
  9. T. J. Sejnowski and P. K. Kienker, 'Learning Symmetry Groups with Hidden Unit: Beyond the perceptron', Physical, 22D, pp.260-275, 1986
  10. Lilly Spirkovska and Max B. Reid, 'Connectivity Strategies for Higher-Order Neural Network Applied to pattern Recognition', IJCNN, 90, San Diego, 1990 https://doi.org/10.1109/IJCNN.1990.137538
  11. 정종수, 홍성찬, '위치 변환 패턴 인식을 위한 다항식 고차 뉴럴 네트워크', 한국정보처리학회 논문지, 제4권, 12호, pp.3063-3068, 1997
  12. 정종수, 하기와라 마사후미, '고차 뉴런을 이용한 Kohohen의 자기 조직화 맵', 대한전기학회 하계학술대회 논문집(D), pp.2656-2659, 7월, 2001
  13. 대한전기학회 하계학술대회 논문집(D) 고차 뉴런을 이용한 Kohohen의 자기 조직화 맵 정종수;하기와라 마사후미