Abstract
This paper presents new algorithms to solve Multi-Objective Markov Decision Processes (MOMDPs). Namely, we present Multi-objective Dynamic Programming variants of Value Iteration such that the values for every state are updated in some heuristic order. The performance of these algorithms is evaluated applying them to benchmark problems with two and three objectives.
Supported by: Plan Propio de Investigación de la Universidad de Málaga - Campus de Excelencia Internacional Andalucía Tech. L. Mandow is supported by project Rhea P18-FR-1081 funded by Junta de Andalucía (co-financed by FEDER funds), Spain.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
References
Barrett, L., Narayanan, S.: Learning all optimal policies with multiple criteria. In: Proceedings of the 25th International Conference on Machine Learning, pp. 41–47 (2008)
Dai, P., Hansen, E.A.: Prioritizing Bellman backups without a priority queue. In: Proceedings of the Seventeenth International Conference on International Conference on Automated Planning and Scheduling, pp. 113–119. AAAI Press (2007)
Drugan, M., Wiering, M., Vamplew, P., Chetty, M.: Special issue on multi-objective reinforcement learning. Neurocomputing 263, 1–2 (2017)
Mandow, L., Pérez de la Cruz, J.L., Pozas, N.: Multi-objective dynamic programming with limited precision. arXiv:2009.08198 (2020)
Mausam, A.K.: Planning with Markov decision processes. An AI perspective. In: Synthesis Lectures on Artificial Intelligence and Machine Learning. Morgan & Claypool (2012)
Roijers, D.M., Whiteson, S.: Multi-objective decision making. In: Synthesis Lectures on Artificial Intelligence and Machine Learning. Morgan & Claypool (2017)
Vamplew, P., Dazeley, R., Berry, A., Issabekov, R., Dekker, E.: Empirical evaluation methods for multiobjective reinforcement learning algorithms. Mach. Learn. 84, 51–80 (2011)
White, D.J.: Multi-objective infinite-horizon discounted Markov decision processes. J. Math. Anal. Appl. 89, 639–647 (1982)
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2021 Springer Nature Switzerland AG
About this paper
Cite this paper
Sedova, E., Mandow, L., Pérez-de-la-Cruz, JL. (2021). Asynchronous Vector Iteration in Multi-objective Markov Decision Processes. In: Alba, E., et al. Advances in Artificial Intelligence. CAEPIA 2021. Lecture Notes in Computer Science(), vol 12882. Springer, Cham. https://doi.org/10.1007/978-3-030-85713-4_13
Download citation
DOI: https://doi.org/10.1007/978-3-030-85713-4_13
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-030-85712-7
Online ISBN: 978-3-030-85713-4
eBook Packages: Computer ScienceComputer Science (R0)