A Robust Principal Component Neural Network

  • Changha Hwang (Associate Professor, Dept, of Statistical Information, Catholic University of Daegu) ;
  • Park, Hyejung (Computer Education Center, Catholic University of Daegu) ;
  • A, Eunyoung-N (Dept. of Statistical Information, Catholic University of Daegu)
  • Published : 2001.12.01


Principal component analysis(PCA) is a multivariate technique falling under the general title of factor analysis. The purpose of PCA is to Identify the dependence structure behind a multivariate stochastic observation In order to obtain a compact description of it. In engineering field PCA is utilized mainly (or data compression and restoration. In this paper we propose a new robust Hebbian algorithm for robust PCA. This algorithm is based on a hyperbolic tangent function due to Hampel ef al.(1989) which is known to be robust in Statistics. We do two experiments to investigate the performance of the new robust Hebbian learning algorithm for robust PCA.



  1. SAS를 이용한 통계자료분석 송문섭;이영조;조신섭;김병천
  2. Communications in Statistics Simulations v.18 Robust Principal Component Ammann,L.
  3. Electronics Letters v.29 no.21 Robust Estimation of Principal Components by using Neural Network Learning Algorithms Cichocki,A.;Unbehauen,R.
  4. Principal Component Neural Networks Diamantaras,k.I.;Kung,S.Y.
  5. Robust Statistics - The Approach Based on Influence Function Hampel,F.R.;Rousseeuw,P.J.;Ronchetti,E.M.;Stahel,W.A.
  6. The korean Communications in Statistics v.6 no.1 A Diagnostic Method in Principal Factor Analysis Kim,M.G.;Jung,K.
  7. Journal of Mathematical Bilogy v.15 A Simplified Neuron Model as a Principal Component Analyzer Oja,E.
  8. Neural Networks v.5 Principal Components, Minor Components, and Linear Neural Networks Oja,E.
  9. Introduction to Data Compression Oja,E.
  10. The 8th Scandinavian Conference on Image Alalysis v.Ⅱ Image Compression by MLP and PCA Neural Networks Wang,L.;Oja,E.
  11. IEEE Transactions on Neural Network v.6 no.1 Robust Principal Component Analysis by Self-Organizing Rules Based on Statistical Physics Approach Xu,L;Yuille,A.L.