English

Fast Online Exact Solutions for Deterministic MDPs with Sparse Rewards

Machine Learning 2018-05-18 v3 Artificial Intelligence Machine Learning

Abstract

Markov Decision Processes (MDPs) are a mathematical framework for modeling sequential decision making under uncertainty. The classical approaches for solving MDPs are well known and have been widely studied, some of which rely on approximation techniques to solve MDPs with large state space and/or action space. However, most of these classical solution approaches and their approximation techniques still take much computation time to converge and usually must be re-computed if the reward function is changed. This paper introduces a novel alternative approach for exactly and efficiently solving deterministic, continuous MDPs with sparse reward sources. When the environment is such that the "distance" between states can be determined in constant time, e.g. grid world, our algorithm offers O(R2×A2×S)O( |R|^2 \times |A|^2 \times |S|), where R|R| is the number of reward sources, A|A| is the number of actions, and S|S| is the number of states. Memory complexity for the algorithm is O(S+R×A)O( |S| + |R| \times |A|). This new approach opens new avenues for boosting computational performance for certain classes of MDPs and is of tremendous value for MDP applications such as robotics and unmanned systems. This paper describes the algorithm and presents numerical experiment results to demonstrate its powerful computational performance. We also provide rigorous mathematical description of the approach.

Keywords

Cite

@article{arxiv.1805.02785,
  title  = {Fast Online Exact Solutions for Deterministic MDPs with Sparse Rewards},
  author = {Joshua R. Bertram and Xuxi Yang and Peng Wei},
  journal= {arXiv preprint arXiv:1805.02785},
  year   = {2018}
}

Comments

Submitted to NIPS 2018; preprint version posted here. 8 pages content, appendices include pseudocode and proof for algorithm

R2 v1 2026-06-23T01:47:51.449Z