- Keywords: Reinforcement Learning, Exploration, Intrinsic Motivation, Sparse Rewards
- TL;DR: A new intrinsic reward signal based on successor features and a novel way to combine extrinsic and intrinsic reward.
- Abstract: Exploration in sparse reward reinforcement learning remains an open challenge. Many state-of-the-art methods use intrinsic motivation to complement the sparse extrinsic reward signal, giving the agent more opportunities to receive feedback during exploration. Commonly these signals are added as bonus rewards, which results in a mixture policy that neither conducts exploration nor task fulfillment resolutely. In this paper, we instead learn separate intrinsic and extrinsic task policies and schedule between these different drives to accelerate exploration and stabilize learning. Moreover, we introduce a new type of intrinsic reward denoted as successor feature control (SFC), which is general and not task-specific. It takes into account statistics over complete trajectories and thus differs from previous methods that only use local information to evaluate intrinsic motivation. We evaluate our proposed scheduled intrinsic drive (SID) agent using three different environments with pure visual inputs: VizDoom, DeepMind Lab and DeepMind Control Suite. The results show a substantially improved exploration efficiency with SFC and the hierarchical usage of the intrinsic drives. A video of our experimental results can be found at https://gofile.io/?c=HpEwTd.
- Code: https://gofile.io/?c=XsePY8