Preference-based Evolutionary Direct Policy Search
Róbert Busa‐Fekete, Balázs Szörényi, Paul Weng, Weiwei Cheng, Eyke Hüllermeier · 2013
Abstract. We present a novel approach to preference-based reinforcement learning, namely a preference-based variant of a direct policy search method based on evolutionary optimization. The core of our approach is a preference-based racing algorithm that selects the best among a given set of candidate policies with high probability. To this end, the algorithm operates on an ordinal preference structure and only uses pairwise comparisons between sample rollouts of the policies. We present first experimental studies showing that our approach performs well in practice. 1