Learning Symbolic Models of Stochastic Domains

TL;DR

This paper introduces a probabilistic relational planning rule framework for learning models of complex, noisy, and nondeterministic domains, demonstrated through experiments in simple and physics-based simulated environments.

Contribution

It presents a novel probabilistic relational rule representation and an effective learning algorithm for modeling stochastic domain dynamics.

Findings

Successfully models noisy, nondeterministic effects

Effective in simple planning and 3D physics domains

Enables agents to learn complex world dynamics

Abstract

In this article, we work towards the goal of developing agents that can learn to act in complex worlds. We develop a probabilistic, relational planning rule representation that compactly models noisy, nondeterministic action effects, and show how such rules can be effectively learned. Through experiments in simple planning domains and a 3D simulated blocks world with realistic physics, we demonstrate that this learning algorithm allows agents to effectively model world dynamics.