An Empirical Study of Reward Structures for Actor-Critic Reinforcement Learning in Air Combat Manoeuvring Simulation

被引：13

作者：

Kurniawan, Budi ^{[1
]}

Vamplew, Peter ^{[1
]}

Papasimeon, Michael ^{[2
]}

Dazeley, Richard ^{[3
]}

Foale, Cameron ^{[1
]}

机构：

[1] Federation Univ, Mt Helen, Vic 3350, Australia

[2] Def Sci & Technol Grp, Fishermans Bend, Vic 3207, Australia

[3] Deakin Univ, Sch Informat Technol, Geelong, Vic 3220, Australia

来源：

AI 2019: ADVANCES IN ARTIFICIAL INTELLIGENCE | 2019年 / 11919卷

关键词：

Reinforcement learning; Actor-critic; Air combat; GAME;

D O I：

10.1007/978-3-030-35288-2_5

中图分类号：

TP18 [人工智能理论];

学科分类号：

081104 ; 0812 ; 0835 ; 1405 ;

摘要：

Reinforcement learning techniques for solving complex problems are resource-intensive and take a long time to converge, prompting a need for methods that encourage faster learning. In this paper we show our successful application of actor-critic reinforcement learning to the air combat simulation domain and how reward structures affect the learning speed to find effective air combat tactics.

引用

页码：54 / 65

页数：12

共 29 条

[1]

Alford R., 2015, P 3 ANN C ADV COGN S

[2] NEURONLIKE ADAPTIVE ELEMENTS THAT CAN SOLVE DIFFICULT LEARNING CONTROL-PROBLEMS [J].

BARTO, AG ;

SUTTON, RS ;

ANDERSON, CW .

IEEE TRANSACTIONS ON SYSTEMS MAN AND CYBERNETICS, 1983, 13 (05) :834-846

[3]

Bengio Y., 2009, P 26 ANN INT C MACH, P41, DOI [DOI 10.1145/1553374.1553380.EVENT-PLACE, 10.1145/1553374.1553380, DOI 10.1145/1553374.15533802,5]

[4]

Fang Jun, 2017, DESTECH T ENG TECHNO

[5]

Floyd MW, 2017, PROCEEDINGS OF THE TWENTY-SIXTH INTERNATIONAL JOINT CONFERENCE ON ARTIFICIAL INTELLIGENCE, P4714

[6]

Grzes M, 2009, LECT NOTES COMPUT SC, V5495, P360, DOI 10.1007/978-3-642-04921-7_37

[7]

Heinze C, 2008, WHITESTEIN SER SOFTW, P113, DOI 10.1007/978-3-7643-8571-2_7

[8]

Lee D, 2013, ADV INTELL SYST, V193, P533

[9]

Liu P., 2017, Communications in Computer and Information Science, P274, DOI DOI 10.1007/978-981-10-6463-024

[10]

Masek Martin, 2018, PRIMA 2018: Principles and Practice of Multi-Agent Systems. 21st International Conference. Proceedings: Lecture Notes in Artificial Intelligence (LNAI 11224), P19, DOI 10.1007/978-3-030-03098-8_2

← 1 2 3 →