UCB Exploration via Q-Ensembles

ICLR 2018 · Richard Y. Chen, Szymon Sidor, Pieter Abbeel, John Schulman ·

We show how an ensemble of $Q^*$-functions can be leveraged for more effective exploration in deep reinforcement learning. We build on well established algorithms from the bandit setting, and adapt them to the $Q$-learning setting. We propose an exploration strategy based on upper-confidence bounds (UCB). Our experiments show significant gains on the Atari benchmark.