Robustness of policies in constrained Markov decision processes

被引：15

作者：

Zadorojniy, A ^{[1
]}

Shwartz, A

机构：

[1] Intel Israel Ltd, MATAM, IL-31015 Haifa, Israel

[2] Technion Israel Inst Technol, Fac Elect Engn, IL-32000 Haifa, Israel

来源：

IEEE TRANSACTIONS ON AUTOMATIC CONTROL | 2006年 / 51卷 / 04期

关键词：

constrained Markov decision process (MDP); discounted cost; Markov decision processes; robustness; sensitivity;

D O I：

10.1109/TAC.2006.872754

中图分类号：

TP [自动化技术、计算机技术];

学科分类号：

0812 ;

摘要：

We consider the optimization of finite-state, finite-action Markov decision processes (MDPs), under constraints. Cost and constraints are discounted. We introduce a new method for investigating the continuity, and a certain type of robustness, of the optimal cost and the optimal policy under changes in the constraints. This method is also applicable for other cost criteria such as finite horizon and infinite horizon average cost.

引用

页码：635 / 638

页数：4

共 50 条

[41] Joint chance-constrained Markov decision processes
V Varagapriya
Vikas Vikram Singh
Abdel Lisser
[J]. Annals of Operations Research, 2023, 322 : 1013 - 1035
[42] Planning using hierarchical constrained Markov decision processes
Seyedshams Feyzabadi
Stefano Carpin
[J]. Autonomous Robots, 2017, 41 : 1589 - 1607
[43] MONOTONE OPTIMAL POLICIES FOR MARKOV DECISION-PROCESSES
SERFOZO, RF
[J]. MATHEMATICAL PROGRAMMING STUDY, 1976, 6 (DEC): : 202 - 215
[44] Learning Policies for Markov Decision Processes in Continuous Spaces
Paternain, Santiago
Bazerque, Juan Andres
Small, Austin
Ribeiro, Alejandro
[J]. 2018 IEEE CONFERENCE ON DECISION AND CONTROL (CDC), 2018, : 4751 - 4758
[45] Finding Safe Zones of Markov Decision Processes Policies
Cohen, Lee
Mansour, Yishay
Moshkovitz, Michal
[J]. ADVANCES IN NEURAL INFORMATION PROCESSING SYSTEMS 36 (NEURIPS 2023), 2023,
[46] Learning Policies for Markov Decision Processes From Data
Hanawal, Manjesh Kumar
Liu, Hao
Zhu, Henghui
Paschalidis, Ioannis Ch.
[J]. IEEE TRANSACTIONS ON AUTOMATIC CONTROL, 2019, 64 (06) : 2298 - 2309
[47] Efficient Policies for Stationary Possibilistic Markov Decision Processes
Ben Amor, Nahla
El Khalfi, Zeineb
Fargier, Helene
Sabaddin, Regis
[J]. SYMBOLIC AND QUANTITATIVE APPROACHES TO REASONING WITH UNCERTAINTY, ECSQARU 2017, 2017, 10369 : 306 - 317
[48] Light robustness in the optimization of Markov decision processes with uncertain parameters
Buchholz, Peter
Scheftelowitsch, Dimitri
[J]. COMPUTERS & OPERATIONS RESEARCH, 2019, 108 : 69 - 81
[49] Sufficiency of Markov Policies for Continuous-Time Jump Markov Decision Processes
Feinberg, Eugene A.
Mandava, Manasa
Shiryaev, Albert N.
[J]. MATHEMATICS OF OPERATIONS RESEARCH, 2022, 47 (02) : 1266 - 1286
[50] Constrained Multiagent Markov Decision Processes: a Taxonomy of Problems and Algorithms
de Nijs, Frits
Walraven, Erwin
de Weerdt, Mathijs M.
Spaan, Matthijs T. J.
[J]. JOURNAL OF ARTIFICIAL INTELLIGENCE RESEARCH, 2021, 70 : 955 - 1001

← 1 2 3 4 5 →