Communications for Statistical Applications and Methods

The Korean Statistical Society (한국통계학회)
권호 표지
DOI QR코드

DOI QR Code

A Method of Determining the Scale Parameter for Robust Supervised Multilayer Perceptrons

  • Park, Ro-Jin (Department of Information & Statistics, Dankook University)
  • Published : 2007.12.31
Download PDF
⟨ Previous Next ⟩

Abstract

Lee, et al. (1999) proposed a unique but universal robust objective function replacing the square objective function for the radial basis function network, and demonstrated some advantages. In this article, the robust objective function in Lee, et al. (1999) is adapted for a multilayer perceptron (MLP). The shape of the robust objective function is formed by the scale parameter. Another method of determining a proper value of that parameter is proposed.

Keywords

References

  1. Andrzej, G. and Dugundji, j. (2003). Fixed Point Theory. Springer-Verlag, New York
  2. Chen, D. S. and Jain, R C. (1994). Robust back propagation learning algorithm for function approximation. IEEE Transactions on Neural Networks, 5, 467-479 https://doi.org/10.1109/72.286917
  3. Freeman, J. A. (1994). Simulating Neural Networks with Mathematica. Addison-Wiley, New York
  4. Hampel, F. R, Ronchetti, E. M., Rousseeuw, P. J. and Stahel, W. A. (1986). Robust Statistics: The Approach Based on Influence Functions. Addison-Wiley, New York
  5. Hecht-Nielsen, R.(1990). Neurocomputing. Addison-Wesley, Reading, MA
  6. Lee, C.-C., Chung, P.-C., Tsai, J.-R, and Chang, C.-I. (1999). Robust radial basis function neural networks. IEEE Transactions on System, Man and Cybernetics, 29, 674-685
  7. Liano, K. (1996). Robust error measure for supervised neural network learning with outliers. IEEE Transactions on Neural Networks, 7, 246-250 https://doi.org/10.1109/72.478411
  8. Moody, J. and Darken, C. (1989). Fast learning in networks for function interpolation. Neural Computation, 3, 281-284
  9. Saha, A., Wu, C. L. and Tang, D. S. (1993). Approximation, dimension reduction and nonconvex optimization using linear superposition of Gaussians. IEEE Transaction on Computers, 42, 1222-1233 https://doi.org/10.1109/12.257708