Skip to main content

Advertisement

Robust Control Methods for On-Line Statistical Learning

Article metrics

  • 585 Accesses

Abstract

The issue of controlling that data processing in an experiment results not affected by the presence of outliers is relevant for statistical control and learning studies. Learning schemes should thus be tested for their capacity of handling outliers in the observed training set so to achieve reliable estimates with respect to the crucial bias and variance aspects. We describe possible ways of endowing neural networks with statistically robust properties by defining feasible error criteria. It is convenient to cast neural nets in state space representations and apply both Kalman filter and stochastic approximation procedures in order to suggest statistically robustified solutions for on-line learning.

Author information

Correspondence to Enrico Capobianco.

Rights and permissions

Reprints and Permissions

About this article

Cite this article

Capobianco, E. Robust Control Methods for On-Line Statistical Learning. EURASIP J. Adv. Signal Process. 2001, 287964 (2001) doi:10.1155/S1110865701000178

Download citation

Keywords

  • artificial learning
  • statistical control algorithms
  • robustness and efficiency of estimators
  • maximum likelihood inference