Learning Action Models: Qualitative Approach

TL;DR

This paper studies how to learn action models in dynamic epistemic logic from observations, establishing criteria for finite and limit identifiability, and proposing specific learning methods for deterministic actions.

Contribution

It introduces a framework for learning action models, analyzes learnability criteria, and develops methods for finite identifiability of deterministic actions.

Findings

Deterministic actions are finitely identifiable.

Non-deterministic actions are identifiable in the limit.

Learning methods resemble update procedures from dynamic epistemic logic.

Abstract

In dynamic epistemic logic, actions are described using action models. In this paper we introduce a framework for studying learnability of action models from observations. We present first results concerning propositional action models. First we check two basic learnability criteria: finite identifiability (conclusively inferring the appropriate action model in finite time) and identifiability in the limit (inconclusive convergence to the right action model). We show that deterministic actions are finitely identifiable, while non-deterministic actions require more learning power-they are identifiable in the limit. We then move on to a particular learning method, which proceeds via restriction of a space of events within a learning-specific action model. This way of learning closely resembles the well-known update method from dynamic epistemic logic. We introduce several different learning methods suited for finite identifiability of particular types of deterministic actions.