Abstract
The learning of agents in a competitive space such as a game is a challenging task. The aim of the proposed research is to improve the reinforcement learning techniques in a competitive multi-agent for the Polo game. First, the video dataset is prepared. Then, the rules of the Polo game are extracted as a class diagram. An architecture is designed for multi-agent team in the Polo game. Therefore, an algorithm is proposed for the temporal difference in the game belief space for improving reward catching. The reward function is implemented in the agent team. Finally, the research improvement is evaluated by increasing 31 units in comparison with previous work. Therefore, competitive learning in the agent team has been improved.
Similar content being viewed by others
References
Ali MZ, Morghem A, Albadarneh J, Al-Gharaibeh R, Suganthan PN, Reynolds RG (2014) Cultural algorithm applied to the evolution of robotic soccer team tactics: a novel perspective, Congress on Evolutionary Computation, IEEE, pp 2180–2187
Asis KD, Hernandez-Garcia JF, Holland GZ, Sutton RS (2018) multi-step reinforcement learning: a unifying algorithm, 31th AAAI Conference on Artificial Intelligence 32:2902–2909
Baglivo A, Ponti FD, Luca DD, Guidazzoli A, Liguori MC (2013) X3D/X3DOM, Blender Game Engine and OSG4WEB: open source visualization for cultural heritage environments. Digital Heritage Int Congress, IEEE 2:711–718
Banerjee B, Davis CE (2017) Multi-agent path finding with persistence conflicts. IEEE Trans Comput Intell AI in Games, IEEE 9:402–409
Beysolow T (2019) Applied reinforcement learning with Python, Book, Springer
Carmel D, Markovitch S (1998) Model-based learning of interaction strategies in multi-agent systems. Taylor & Francis Ltd 10:309–332
Castelfranchi C, Lesperance Y (2000) Intelligent Agent VII–Agent Theories, Architecture and Languages, Springer
Chen B, Zhang A, Cao L (2014) Autonomous intelligent decision-making system based on Bayesian SOM. Neural Netw Robot Soccer, Elsevier, Neurocomputing 128:447–458
Collazo MN, Cotta C, Fernandez-Leiva AJ (2014) Virtual player design using self-learning via competitive coevolutionary algorithms, Springer. Nat Comput 31:131–144
Collazo MN, Porras CC, Fern’Andez-Leiva AJ (2016) Competitive algorithms for co-evolving both game content and AI a case of study: planet wars. IEEE Transactions on Computational Intelligence and AI in Games 8(4):325–337
Covaci A, Ghinea G, Huang CH, Shih J (2018) Multisensory came-lessons learnt from olfactory enhancement of a digital board game, Springer. Multimed Tools Appl 77:21245–21263
Danny W (2010) Architecture-based design of multi-agent systems. Springer
Ding S, Du W, Zhao X, Wang L, Jia W (2019) A new asynchronous reinforcement learning algorithm based on improved parallel PSO, Springer. Appl Intell 49:4211–4222
Duan Y, Cui BX, Xu XH (2011) A multi-agent reinforcement learning approach to robot soccer, Springer. Artif Intell 36:193–211
Fernando TG, Luis Javier Garcia V, Ana Lucila SO, Kim TH (2019) A comparison of learning methods over raw data: forecasting cab services market share in New York city, Springer. Multimed Tools Appl 78:29783–29804
Guimaraes M, Santos P, Jhala A (2017) Prom week meet Skyrim: developing a social agent architecture in a commercial game. ACM, 17th international conference on autonomous agent and multi-agent system, pp 1562–1564
Hagelbäck J (2016) Hybrid path finding in StarCraft. IEEE, Transactions on Computational Intelligence and AI in Games 38:319–324
Hajuk M, Sukop M, Haun M (2019) Cognitive Multi-agent Systems, Book, Springer
Hübner JF, Bordini RH (2010) Using agent-and organization-oriented programming to develop a team of agents for a competitive game. Springer, Science Business Media 59:351–372
Husseinzadeh Kashan A, Karimi B (2010) A new algorithm for constrained optimization inspired by the sport leagues championships. IEEE. Congress on evolutionary computation. https://ieeexplore.ieee.org/document/5586364
Kamalapurkar R, Walters P, Rosenfeld DW (2018) Reinforcement learning for optimal feedback control. Springer
Kim JH, Vadakkepat P (2000) Multi-agent systems: a survey from the robot-soccer perspective. Intell Autom Soft Comput 6:3–17
Kobti Z, Sharma S (2007) A multi-agent architecture for game playing. Computational intelligence and games. Proceedings of the 2007 IEEE Symposium on Computational Intelligence and Games. https://ieeexplore.ieee.org/document/4219054
Koseler K, Stephan M (2018) Machine learning applications in baseball: a systematic literature review, applied artificial intelligence, applied artificial intelligence. Taylor & Francis, pp 745–763
Laffay HA (2011) Polo in the United States: a history. Book. MC Far Land & Company Inc
Laffay HA (2014) Polo in Argentina: a history. MC Far Land & Company Inc
Horace A Laffaye (2009) The evaluation of polo. McFarlane & Company Inc
Lee JM, Lee BJ, Kim KE (2020) Reinforcement learning for control with multiple frequencies. NeurIPS. 34th conference on neural information processing systems, pp 1–11
Leng J, Lakhmi J (2009) Experimental analysis of eligibility traces strategies in temporal difference learning. Knowl Eng Soft Data Paradigms 1:26–39
Leng J, Lim CP (2011) Reinforcement learning of competitive and cooperative skills in soccer agents. Elsevier. Appl Soft Comput 11:1353–1363
Leng J, Fyfe C, Lakhmi J (2007) Reinforcement learning of competitive skills with soccer agents. Springer. Proc 11th Knowledge-Based Intell Inf Eng Syst 4692:572–579
Marinheiro J, Cardoso HL (2017) A generic agent architecture for cooperative multi-agent games. ICAART. Proceedings of the 9th International Conference on Agents and Artificial Intelligence 2:107-118
Masoumi B, Meybodi MR (2010) Learning automata based multi-agent system algorithms for finding optimal policies in Markov games. J Control 14:137–152
Masoumi B, Meybodi MR (2011) Speeding up learning automata based multi-agent the concept of Stigmergy and entropy. Elsevier. Expert Systems with Applications 38:8105–8118
Masoumi B, Meybodi MR, Abtahi F (2012) Learning automata based algorithms for finding optimal policies in fully cooperative Markov games. PRZEGLĄD ELEKTROTECHNICZNY 8:280–289
Mattiassi ADA (2019) Fighting the game. Command systems and player-avatar interaction in fighting games in a social cognitive neuroscience framework. Springer. Multimedia Tools and Applications 78:13565–13591
Mourao A, Magalhaes J (2013) Competitive affective gaming: winning with smile, proceedings of the 21st ACM international conference on multimedia, ACM, pp 83–91
Nandy A, Biswas M (2018) Reinforcement learning. Springer
Nash A, Koenig S (2013) Any-angle path planning. American Association for Artificial Intelligence. AI Magazine 34:85–107
Parag C, Pendharkar (2012) Game theoretical applications for multi-agent systems. Elsevier. Expert Systems with Applications 39:273–279
Pelechano N, Fuentes C (2016) Hierarchical path-finding for navigation mesh. Elsevier. Computers Graphics 59:68–78
Polk S, Oommen BJ (2018) Novel threat-based AI strategies that incorporate adaptive data structures for multi-player board game. Springer. Applied Intelligence 48:1893–1911
Polo Federation of Iran site: (n.d.) http://iranPolo.org/?page_id=1521#lightbox[gallery-1]/1/
Rabin S (2002) AI programming WISDOM. Charles River media. April 3
Scheepers C, Engelbrecht AP (2014) Competitive co-evolutionary training of simple soccer agents from zero knowledge. IEEE. Congress on Evolutionary Computation. https://ieeexplore.ieee.org/document/6900236
Scheepers C, Engelbrecht AP (2014) Training multi-agent teams from zero knowledge with the competitive co-evolutionary team-based particle swarm optimizer. Springer. Soft Compute 20:607–620
Scheepers C, Engelbrecht AP (2014) Analysis of stagnation behavior of competitive co-evolutionary trained neuro-controller. IEEE. Symposium on Swarm Intelligence. https://ieeexplore.ieee.org/document/7011795
Sewak M (2019) Deep reinforcement learning. Springer
Stone P, Veloso M (1998) Layered approach to learning client behaviors in the Robocup. Taylor & Francis. Applied Artificial Intelligence 12:165–188
Sun P, Hu Y, Lan J, Tian L, Chen M (2019) TIDE: time-relevant deep reinforcement learning for routing optimization. Elsevier. Future Generation Computer Systems 99:401–409
Tomaz L.B.P, Julia R.M.S, Duarte V.A (2017) A multi-agent player system composed by expert agents in specific game stages operating in high performance environment. Springer. Applied Intelligence 48:1–22
Weyns D, Mascarsdi V, Ricci A (2019) Engineering multi-agent systems. Springer
Wooldridge M (2002) An introduction to multi-agent systems. John Wiley& Sons. August
Wooldridge M, Mller J, Tambe M (1997) Intelligent Agent II –Agent Theories. Architecture and Languages. Springer
Yu FR, He Y (2019) Deep reinforcement learning for wireless networks. Springer
Yuan Y, Yu Z.L, Gu Z, Deng X, Li X (2019) A novel multi-step reinforcement learning method for solving reward hacking. Springer. Applied Intelligence 49:2878–2888
Author information
Authors and Affiliations
Corresponding author
Additional information
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
About this article
Cite this article
Movahedi, Z., Bastanfard, A. Toward competitive multi-agents in Polo game based on reinforcement learning. Multimed Tools Appl 80, 26773–26793 (2021). https://doi.org/10.1007/s11042-021-10968-z
Received:
Revised:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11042-021-10968-z