Skip to main content

June 5, 2017

UCB exploration via Q-ensembles

Ucb Exploration Via Q Ensembles


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.