Continual Reinforcement Learning in 3D Non-stationary Environments
This work addresses the problem of adapting reinforcement learning agents to unpredictable real-world changes, though it is incremental as it builds on existing benchmarks and methods.
The authors tackled the challenge of reinforcement learning in high-dimensional, non-stationary environments by proposing CRLMaze, a new benchmark based on ViZDoom, and introduced a model-free continual reinforcement learning strategy that achieved competitive results against four baselines without needing extra supervision or prior environmental data.
High-dimensional always-changing environments constitute a hard challenge for current reinforcement learning techniques. Artificial agents, nowadays, are often trained off-line in very static and controlled conditions in simulation such that training observations can be thought as sampled i.i.d. from the entire observations space. However, in real world settings, the environment is often non-stationary and subject to unpredictable, frequent changes. In this paper we propose and openly release CRLMaze, a new benchmark for learning continually through reinforcement in a complex 3D non-stationary task based on ViZDoom and subject to several environmental changes. Then, we introduce an end-to-end model-free continual reinforcement learning strategy showing competitive results with respect to four different baselines and not requiring any access to additional supervised signals, previously encountered environmental conditions or observations.