Abstract
Markov Decision Processes (MDPs) are fundamental in computational decision-making frameworks, requiring efficient computational methods for their resolution. Traditionally, methods such as matrix splitting have been used to decompose the Bellman equations for iterative solutions. However, the Krylov Subspace Method (KSM) presents a more efficient alternative, leveraging dynamic iterative techniques that significantly outperform static ones. This paper introduces the Enhanced Iterative Subspace Method (EISM), detailing its conceptual motivations, practical implementations, and superiority through various computational experiments.
Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 License.
Recommended Citation
"Optimized Iterative Methods for Markov Decision Processes: A Subspace Approach", Technical Disclosure Commons, (April 16, 2024)
https://www.tdcommons.org/dpubs_series/6881