Learning to Act and Observe in Partially Observable Domains
This work addresses the challenge of learning in partially observable domains for AI agents, building incrementally on prior research in fully observable settings.
The paper tackles the problem of a learning agent acquiring knowledge about both observable aspects and action effects in partially observable environments, presenting algorithms that use dynamic epistemic logic to represent learned domain information and characterize the observational requirements for achieving different levels of domain knowledge.
We consider a learning agent in a partially observable environment, with which the agent has never interacted before, and about which it learns both what it can observe and how its actions affect the environment. The agent can learn about this domain from experience gathered by taking actions in the domain and observing their results. We present learning algorithms capable of learning as much as possible (in a well-defined sense) both about what is directly observable and about what actions do in the domain, given the learner's observational constraints. We differentiate the level of domain knowledge attained by each algorithm, and characterize the type of observations required to reach it. The algorithms use dynamic epistemic logic (DEL) to represent the learned domain information symbolically. Our work continues that of Bolander and Gierasimczuk (2015), which developed DEL-based learning algorithms based to learn domain information in fully observable domains.