On Avoidance Learning with Partial Observability
This addresses avoidance learning for agents in partially observable environments, but it appears incremental as it builds on existing frameworks with a new algorithm.
The paper tackles the problem of agents learning to avoid aversive signals under partial observability and non-determinism, resulting in a parameter-free algorithm called A-learning that achieves fixpoint convergence.
We study a framework where agents have to avoid aversive signals. The agents are given only partial information, in the form of features that are projections of task states. Additionally, the agents have to cope with non-determinism, defined as unpredictability on the way that actions are executed. The goal of each agent is to define its behavior based on feature-action pairs that reliably avoid aversive signals. We study a learning algorithm, called A-learning, that exhibits fixpoint convergence, where the belief of the allowed feature-action pairs eventually becomes fixed. A-learning is parameter-free and easy to implement.