Towards a Unified Framework for Sequential Decision Making

TL;DR

This paper proposes a unified probabilistic framework for Sequential Decision Making that integrates Automated Planning and Reinforcement Learning, enabling better understanding and comparison of different methods.

Contribution

It introduces a general SDM framework applicable to various methods, along with algorithms and formulas for evaluation and comparison.

Findings

A unified framework for SDM encompassing AP and RL.

Algorithms for empirical evaluation of SDM methods.

Formulas for analyzing properties of SDM tasks.

Abstract

In recent years, the integration of Automated Planning (AP) and Reinforcement Learning (RL) has seen a surge of interest. To perform this integration, a general framework for Sequential Decision Making (SDM) would prove immensely useful, as it would help us understand how AP and RL fit together. In this preliminary work, we attempt to provide such a framework, suitable for any method ranging from Classical Planning to Deep RL, by drawing on concepts from Probability Theory and Bayesian inference. We formulate an SDM task as a set of training and test Markov Decision Processes (MDPs), to account for generalization. We provide a general algorithm for SDM which we hypothesize every SDM method is based on. According to it, every SDM algorithm can be seen as a procedure that iteratively improves its solution estimate by leveraging the task knowledge available. Finally, we derive a set of formulas and algorithms for calculating interesting properties of SDM tasks and methods, which make possible their empirical evaluation and comparison.