Efficient Asymptotic Approximation in Temporal Difference Learning

Efficient Asymptotic Approximation in Temporal Difference Learning
European Conference on Artificial Intelligence ECAI'2000 ( gzipped Postscript - 78383 KB )

Abstract: We propose in this paper an asymptotic approximation of online TD(lambda) with accumulating eligibility trace, called ATD(lambda). We then use the Ordinary Differential Equation (ODE) method to analyse ATD(lambda) and to optimize the choice of the lambda parameter and the learning stepsize, and we introduce ATD, a new efficient temporal difference learning algorithm.