DQN with model-based exploration: efficient learning on environments with sparse rewards

Open in new window