Deep Reinforcement Learning with Double Q-learning

Abstract

This paper introduces and evaluates the idea described in “Deep Reinforcement Learning with Double Q-learning”, and reports empirical results that helped shape subsequent work in reinforcement-learning, dqn.