Session:

Neural Networks and Fuzzy Systems
Tuesday, March 02, 2004, 17.35 – 17.55

Session Chair:

Peter Anderson

Paper Title:

A Derivative-Free Kalman Filter for Parameter Estimation of Recurrent Neural Networks and Its Applications to Nonlinear Channel Equalization

Author(s):

J. Choi, University of Ottawa, Canada
M. Bouchard, University of Ottawa, Canada
T. Hin Yeap, University of Ottawa, Canada
O. Kwon, School of Electronics and Information Engineering, Kunsan National University, Kunsan, Korea

Abstract:

Recurrent neural networks (RNNs) trained with gradient-based algorithms such as real-time recurrent learning or back-propagation through time have a drawback of slow convergence rate. These algorithms also need the derivative calculation which is not trivialized in through the error back-propagation process. In this paper, a derivative-free Kalman filter, so called the unscented Kalman filter (UKF), for training a fully connected RNN is presented in a state-space formulation of the system. The UKF algorithm makes the RNN have fast convergence speed and good tracking performance without the derivative computation. Through experiments of nonlinear channel equalization, the performance of the RNN with the UKF is evaluated.