Reinforcement learning with rare significant events: direct policy search vs. gradient policy search.

Paul Ecoffet, Nicolas Fontbonne, Jean-Baptiste Andr, Nicolas Bredche

Browse the full GECCO paper archive.