1 paper · 1 filter
Manuel Loth, Philippe Preux
This paper addresses the issue of policy evaluation in Markov Decision Processes, using linear function approximation. It provides a unified view of algorithms such as TD(lambda),…