Reactive Exploration to Cope with Non-Stationarity in Lifelong Reinforcement Learning
This work addresses the challenge of lifelong learning for AI agents in dynamic, real-world settings, though it appears incremental as it builds on existing policy-gradient methods.
The paper tackles the problem of non-stationarity in lifelong reinforcement learning, where agents face continual domain shifts, and shows that policy-gradient methods combined with Reactive Exploration adapt more quickly than Q-learning, achieving good results in such environments.
In lifelong learning, an agent learns throughout its entire life without resets, in a constantly changing environment, as we humans do. Consequently, lifelong learning comes with a plethora of research problems such as continual domain shifts, which result in non-stationary rewards and environment dynamics. These non-stationarities are difficult to detect and cope with due to their continuous nature. Therefore, exploration strategies and learning methods are required that are capable of tracking the steady domain shifts, and adapting to them. We propose Reactive Exploration to track and react to continual domain shifts in lifelong reinforcement learning, and to update the policy correspondingly. To this end, we conduct experiments in order to investigate different exploration strategies. We empirically show that representatives of the policy-gradient family are better suited for lifelong learning, as they adapt more quickly to distribution shifts than Q-learning. Thereby, policy-gradient methods profit the most from Reactive Exploration and show good results in lifelong learning with continual domain shifts. Our code is available at: https://github.com/ml-jku/reactive-exploration.