Minimizing the Fisher Information of the Error in Supervised Adative Filter Training

In this paper, we propose minimizing the Fisher information of the error in supervised training of linear and nonlinear adaptive filters. Fisher information considers the local structure of the error probability distribution and therefore, it is expected to result in more robust solutions compared to other statistics such as minimum mean-square-error or minimum-error-entropy. A gradient-based training algorithm based on a nonparametric estimator of Fisher information is presented and the performances of the three mentioned optimization criteria is compared using Monte Carlo simulations.