DOI QR코드

DOI QR Code

An Improved Robust Fuzzy Principal Component Analysis

잡음 민감성이 개선된 퍼지 주성분 분석

  • Received : 2010.01.26
  • Accepted : 2010.02.12
  • Published : 2010.05.31

Abstract

Principal component analysis (PCA) is a well-known method for dimension reduction while maintaining most of the variation in data. Although PCA has been applied to many areas successfully, it is sensitive to outliers. Several variants of PCA have been proposed to resolve the problem and, among the variants, robust fuzzy PCA (RF-PCA) demonstrated promising results. RF-PCA uses fuzzy memberships to reduce the noise sensitivity. However, there are also problems in RF-PCA and the convergence property is one of them. RF-PCA uses two different objective functions to update memberships and principal components, which is the main reason of the lack of convergence property. The difference between two functions also slows the convergence and deteriorates the solutions of RF-PCA. In this paper, a variant of RF-PCA, called RF-PCA2, is proposed. RF-PCA2 uses an integrated objective function both for memberships and principal components. By using alternating optimization, RF-PCA2 is guaranteed to converge on a local optimum. Furthermore, RF-PCA2 converges faster than RF-PCA and the solutions found are more similar to the desired solutions than those of RF-PCA. Experimental results also support this.

주성분 분석(PCA)은 데이터의 차원을 줄이면서 최대의 데이터 변이를 보존하는 기법으로 차원 축소나 피처 추출을 위해 널리 사용되고 있다. 하지만 PCA는 잡음에 민감한 단점이 있으며, 이러한 잡음 민감성을 해결하기 위해 여러 가지 PCA 변형이 제안되었다. 그 중 robust fuzzy PCA(RF-PCA)는 퍼지 소속도를 사용하여 잡음의 영향을 효과적으로 줄일 수 있음이 입증되었다. 하지만 RF-PCA 역시 몇 가지 문제점이 있고, 수렴성이 그 중 하나이다. RF-PCA는 소속도와 주성분을 갱신할 때 서로 다른 목적 함수를 사용하므로 수렴 속도가 느리고 구해지는 해가 국부 최적 해임을 보장하지 않는다. 이 논문에서는 RF-PCA의 문제점을 해결하기 위해 하나의 목적 함수를 이용해 소속도와 주성분을 갱신할 수 있는 방법을 제안한다. 제안한 방법, RF-PCA2는 반복 최적화를 이용함으로써 국부 최적해에 수렴함을 보장하며, RF-PCA에 비해 빠른 수렴 속도를 가지고, 잡음 민감성이 줄어든다. 이러한 사실들은 실험 결과를 통해 확인할 수 있다.

Keywords

References

  1. I. T. Jolliffe, Principal Component Analysis, 2nd Edition, Springer, 2002.
  2. P. Rousseeuw, "Multivariate estimation with high breakdown point," Mathematical Statistics and Applications B, pp. 283-297, 1985.
  3. C.-D. Lu, T.-Y. Zhang, X.-Z. Du, and C.-P. Li, "A robust kernel PCA algorithm," Proceedings of the 3rd International Conference on Machine Learning and Cybernetics, pp. 3084-3087, 2004.
  4. C. Lu, T. Zhang, R. Zhang, and C. Zhang, "Adaptive robust kernel PCA algorithm," Proceedings of the IEEE International Conference on Acoustics, Speech, and Signal Processing, pp. VI621-624, 2003.
  5. G. Heo, P. Gader, and H. Frigui, "RKF-PCA: Robust kernel fuzzy PCA," Neural Networks Vol. 22, No. 5-6, pp. 642-650, 2009. https://doi.org/10.1016/j.neunet.2009.06.013
  6. T.-N. Yang and S.-D. Wang, "Fuzzy auto-associative neural networks for principal component extraction of noisy data," IEEE Transaction on Neural Networks, Vol. 11, No. 3, pp. 808-810, 2000 https://doi.org/10.1109/72.846752
  7. T. R. Cundari, C. Sarbu, and H. F. Pop, "Robust fuzzy principal component analysis (FPCA). A comparative study concerning interaction of carbonhydrogen bonds with molybdenum-oxo bonds," Journal of Chemical Information and Computer Sciences, Vol. 42, No. 6, pp. 1363-1369, 2002. https://doi.org/10.1021/ci025524s
  8. J. C. Bezdekand R. J. Hathaway, "Convergence of alternating optimization," Neural, Parallel and Scientific Computations, Vol. 11, No. 4, pp. 351-368, 2003.
  9. G. Heo and P. Gader, "Fuzzy SVM for noisy data: A robust membership calculation method," Proceedings of the 2009 IEEE International Conference on Fuzzy Systems, pp. 431-436, 2009.
  10. H. Ichihashi, K. Honda, and N. Tani, "Gaussian mixture PDF approximation and fuzzy c-means clustering with entropy regularization," Proceedings of the 4th Asian Symposium, pp. 217-221, 2000.
  11. P. J. Huber, Robust Statistics, Wiley-Interscience, 1981.

Cited by

  1. 잡음 민감성이 개선된 변형 퍼지 주성분 분석 기법 vol.16, pp.2, 2010, https://doi.org/10.9708/jksci.2011.16.2.025