Centroid Neural Network with Bhattacharyya Kernel

Bhattacharyya 커널을 적용한 Centroid Neural Network

  • 이송재 ((주) LG. Phillips LCD) ;
  • 박동철 (명지대학교 정보공학과 지능컴퓨팅 연구실)
  • Published : 2007.09.30

Abstract

A clustering algorithm for Gaussian Probability Distribution Function (GPDF) data called Centroid Neural Network with a Bhattacharyya Kernel (BK-CNN) is proposed in this paper. The proposed BK-CNN is based on the unsupervised competitive Centroid Neural Network (CNN) and employs a kernel method for data projection. The kernel method adopted in the proposed BK-CNN is used to project data from the low dimensional input feature space into higher dimensional feature space so as the nonlinear problems associated with input space can be solved linearly in the feature space. In order to cluster the GPDF data, the Bhattacharyya kernel is used to measure the distance between two probability distributions for data projection. With the incorporation of the kernel method, the proposed BK-CNN is capable of dealing with nonlinear separation boundaries and can successfully allocate more code vector in the region that GPDF data are densely distributed. When applied to GPDF data in an image classification probleml, the experiment results show that the proposed BK-CNN algorithm gives 1.7%-4.3% improvements in average classification accuracy over other conventional algorithm such as k-means, Self-Organizing Map (SOM) and CNN algorithms with a Bhattacharyya distance, classed as Bk-Means, B-SOM, B-CNN algorithms.

본 논문은 가우시안 확률분포함수 (Gaussian Probability Distribution Function) 데이터 군집화를 위해 중심신경망 (Centroid Neural Network, CNN)에 Bhattacharyya 커널을 적용한 군집화 알고리즘 (Bhattacharyya Kernel based CNN, BK-CNN)을 제안한다. 제안된 BK-CNN은 무감독 알고리즘인 중심신경망을 기반으로 하고 있으며, 커널 방법을 이용하여 데이터를 특징공간에서 투영한다. 입력공간의 비선형 문제를 선형적으로 해결하기 위해 제안한 커널 방법인데, 확률분포 사이의 거리측정을 위해 Bhattacharyya 거리를 이용한 커널방법을 사용하였다. 제안된 BK-CNN을 영상데이터 분류의 문제에 적용했을 때, 제안된 BK-CNN 알고리즘이 Bhattacharyya 커널을 적용한 k-means, 자기조직지도(Self-Organizing Map)와 중심 신경망등의 기존 알고리즘보다 1.7% - 4.3%의 평균 분류정확도 향상을 가져옴을 확인할 수 있었다.

Keywords

References

  1. Hartigan, J., 'Clustering Algorithms,' New York, Wiley, 1975
  2. Kohonen, T., 'The Self-Organizing Map,' Processing of the IEEE, vol. 78, pp, 1464-1480, Sept. 1990
  3. Park, D. C., ,'Centroid Neural Network,' IEEE Tr. on Neural Networks, vol. 11, pp520-528, Mar. 2000 https://doi.org/10.1109/72.839021
  4. Yin, H., Allinson, N. M., 'Bayesian self-organizing map for Gaussian mixture,' IEE Processings Vison Image and Signal Processing, vol. 148, pp.234-240, Aug. 2001 https://doi.org/10.1049/ip-vis:20010378
  5. Gokcay, E., Principe, J.C., 'Information theoretic clustering,' IEEE Tr. on Pattern Analysis and Machine Intelligence, vol. 24, pp. 158-171, Feb. 2002 https://doi.org/10.1109/34.982897
  6. Park, D. C., Oh, H.K., Suk, M. S.,'Clustering of Gaussian Probability Density Functions Using Centroid Neural Network,'IEE Electronic Letters, vol. 49, pp. 381- 382, Feb. 2003
  7. Girolami, M.,'Mercer Kernel-Based Clustering in Feature Space,'IEEE Tr. on Neural Networks, vol. 13, pp. 780-784, May. 2002 https://doi.org/10.1109/TNN.2002.1000150
  8. Cristianini, N., Shawe-Taylor, J., 'An Introduction to Support Vector Machine,' Cambridge, Cambridge Univ., 2000
  9. Chen, S., Zhang, D.,'Robust Image Segmentation using FCM with Spatial Constraints Based on New Kernel-Induced Distance Measure,'IEEE Tr. On Systems Man and Cybernetics, vol. 43, pp. 1907- 1916, Aug. 2004
  10. Jebra, T., Kondor,'Bhattacharyya and Expected Likelihood Kernels,'Proc. COLT, 2003
  11. Muller, K.R., Mika, S., Ratsch, G., Ratsch, G., Tsuda, K., Scholkopf, B., 'An Introduction to Kernel-Based Learning Algorithms,'IEEE Tr. on Neural Networks, vol. 12, pp.181- 201, Mar. 2001 https://doi.org/10.1109/72.914517
  12. Chen, J.H., Chen, C.S., 'Fuzzy Kernel Perceptron,'IEEE Tr. on Neural Networks, vol. 13, pp. 1364- 1373, Nov. 2002 https://doi.org/10.1109/TNN.2002.804311