Abstract
Many real-world multi-agent interactions consider multiple distinct criteria, i.e. the payoffs are multi-objective in nature. However, the same multi-objective payoff vector may lead to different utilities for each participant. Therefore, it is essential for an agent to learn about the behaviour of other agents in the system. In this work, we present the first study of the effects of such opponent modelling on multi-objective multi-agent interactions with nonlinear utilities. Specifically, we consider two-player multi-objective normal form games with nonlinear utility functions under the scalarised expected returns optimisation criterion. We contribute novel actor-critic and policy gradient formulations to allow reinforcement learning of mixed strategies in this setting, along with extensions that incorporate opponent policy reconstruction and learning with opponent learning awareness (i.e. learning while considering the impact of one’s policy when anticipating the opponent’s learning step). Empirical results in five different MONFGs demonstrate that opponent learning awareness and modelling can drastically alter the learning dynamics in this setting. When equilibria are present, opponent modelling can confer significant benefits on agents that implement it. When there are no Nash equilibria, opponent learning awareness and modelling allows agents to still converge to meaningful solutions that approximate equilibria.
Similar content being viewed by others
Notes
References
Albrecht SV, Stone P (2018) Autonomous agents modelling other agents: A comprehensive survey and open problems. Artif Intell 258:66–95
Billingsley P (2008) Probability and measure. Wiley, New York
Blackwell D et al (1956) An analog of the minimax theorem for vector payoffs. Pacific J Math 6(1):1–8
Bonilla E.V, Chai K.M, Williams C (2008) Multi-task gaussian process prediction. Advances in neural information processing systems, 153–160
Borm P, Tijs S, Van Den Aarssen J (1988) Pareto equilibria in multiobjective games. Methods Op Res 60:302–312
Borm P, Vermeulen D, Voorneveld M (2003) The structure of the set of equilibria for two person multicriteria games. Eur J Op Res 148(3):480–493
Chajewska U, Koller D (2000) Utilities as random variables: Density estimation and structure discovery. In Proceedings of the Sixteenth conference on Uncertainty in artificial intelligence, pp 63–71. Morgan Kaufmann Publishers Inc
Chajewska U, Koller D, Ormoneit D (2001) Learning an agent’s utility function by observing behavior. In Proceedings of the Eighteenth International Conference on Machine Learning, pp 35–42
Chajewska U, Koller D, Parr R (2000) Making rational decisions using adaptive utility elicitation. In: AAAI/IAAI, pp. 363–369
Chu W, Ghahramani Z (2005) Preference learning with gaussian processes. In Proceedings of the 22nd international conference on Machine learning, pp 137–144. ACM
Claus C (1998) Boutilier C (1998) The dynamics of reinforcement learning in cooperative multiagent systems. AAAI/IAAI 746–752:2
Foerster J, Chen R.Y, Al-Shedivat M, Whiteson S, Abbeel P, Mordatch I (2018) Learning with opponent-learning awareness. In Proceedings of the 17th International Conference on Autonomous Agents and MultiAgent Systems, pp 122–130
Foerster J, Farquhar G, Al-Shedivat M, Rocktäschel T, Xing E, Whiteson S (2018) Dice: The infinitely differentiable monte carlo estimator. In International Conference on Machine Learning, pp 1529–1538
Fudenberg D, Drew F, Levine DK, Levine DK (1998) The theory of learning in games, vol 2. MIT press, Cambridge
Guo S, Sanner S, Bonilla EV (2010) Gaussian process preference elicitation. Adv Neural Inf Process Syst 23:262–270
Hayes C.F, Rădulescu R, Bargiacchi E, Källström J, Macfarlane M, Reymond M, Verstraeten T, Zintgraf L.M, Dazeley R, Heintz F, et al (2021) A practical guide to multi-objective reinforcement learning and planning. arXiv preprint arXiv:2103.09568
He H, Boyd-Graber J, Kwok K, Daumé III H (2016) Opponent modeling in deep reinforcement learning. In International Conference on Machine Learning, pp 1804–1813
Knegt SJ, Drugan MM, Wiering MA (2018) Opponent modelling in the game of tron using reinforcement learning. ICAART 2:29–40
Lowe R, Wu YI, Tamar A, Harb J, Pieter Abbeel O, Mordatch I (2017) Multi-agent actor-critic for mixed cooperative-competitive environments. Adv Neural Inf Process Syst 30:6379–6390
Lozovanu D, Solomon D, Zelikovsky A (2005) Multiobjective games and determining pareto-nash equilibria. Buletinul Academiei de Ştiinţe a Republicii Moldova. Matematica 3:115–122
Mannion P, Devlin S, Duggan J, Howley E (2018) Reward shaping for knowledge-based multi-objective multi-agent reinforcement learning. Knowl Eng Rev 33:e23
Nash J (1951) Non-cooperative games. Annals Math 54(2):286–295
Ng A.Y, Russell S.J, et al (2000) Algorithms for inverse reinforcement learning. In Proceedings of the Seventeenth International Conference on Machine Learning, vol. 1, p 2
Nisan N, Roughgarden T, Tardos E, Vazirani VV (2007) Algorithmic game theory. Cambridge University Press, Cambridge
Paszke A, Gross S, Massa F, Lerer A, Bradbury J, Chanan G, Killeen T, Lin Z, Gimelshein N, Antiga L, Desmaison A, Kopf A, Yang E, DeVito Z, Raison M, Tejani A, Chilamkurthy S, Steiner B, Fang L, Bai J, Chintala S (2019) PyTorch: An Imperative Style, High-Performance Deep Learning Library. In: H. Wallach, H. Larochelle, A. Beygelzimer, F. d’Alché Buc, E. Fox, R. Garnett (eds.) Advances in Neural Information Processing Systems, vol. 32, pp. 8026–8037. Curran Associates, Inc. https://proceedings.neurips.cc/paper/2019/file/bdbca288fee7f92f2bfa9f7012727740-Paper.pdf
Perolat J, Leibo J.Z, Zambaldi V, Beattie C, Tuyls K, Graepel T (2017) A multi-agent reinforcement learning model of common-pool resource appropriation. In Proceedings of the 31st International Conference on Neural Information Processing Systems, pp 3646–3655
Raileanu R, Denton E, Szlam A, Fergus R Modeling others using oneself in multi-agent reinforcement learning. In International Conference on Machine Learning (ICML)
Rasmussen CE, Kuss M (2003) Gaussian processes in reinforcement learning. Adv Neural Inf Process Syst 16:751–758
Roijers D.M, Steckelmacher D, Nowé A (2018) Multi-objective reinforcement learning for the expected utility of the return. In Proceedings of the Adaptive and Learning Agents workshop at FAIM
Roijers DM, Vamplew P, Whiteson S, Dazeley R (2013) A survey of multi-objective sequential decision-making. J Artif Intell Res 48:67–113
Roijers DM, Whiteson S (2017) Multi-objective decision making. Synth Lect Artif Intell Mach Learn 11(1):1–129
Rădulescu R, Mannion P, Roijers DM, Nowé A (2020) Multi-objective multi-agent decision making: a utility-based analysis and survey. Auton Agents Multi-Agent Syst 34(1):1–52. https://doi.org/10.1007/s10458-019-09433-x
Rădulescu R, Mannion P, Zhang Y, Roijers DM, Nowé A (2020) A utility-based analysis of equilibria in multi-objective normal-form games. Knowl Eng Rev 35:e32. https://doi.org/10.1017/S0269888920000351
Shapley LS, Rigby FD (1959) Equilibrium points in games with vector payoffs. Naval Res Logist Quart 6(1):57–61
Shen Y, Wu Y, Chen G, Van Grinsven HJ, Wang X, Gu B, Lou X (2017) Non-linear increase of respiratory diseases and their costs under severe air pollution. Environ Pollut 224:631–637
Sutton R.S, Barto A.G (2018) Reinforcement Learning: An Introduction, second edn. The MIT Press. http://incompleteideas.net/book/the-book-2nd.html
Uther W, Veloso M (1997) Adversarial reinforcement learning. Technical report, Carnegie Mellon University, USA (Unpublished)
Voorneveld M, Vermeulen D, Borm P (1999) Axiomatizations of pareto equilibria in multicriteria games. Games Econ Behav 28(1):146–154
Wang KA, Pleiss G, Gardner JR, Tyree S, Weinberger KQ, Wilson AG (2019) Exact gaussian processes on a million data points. Advances in Neural Information Processing Systems 32
Williams RJ (1992) Simple statistical gradient-following algorithms for connectionist reinforcement learning. Mach Learn 8(3–4):229–256. https://doi.org/10.1007/BF00992696
Wilson A, Nickisch H (2015) Kernel interpolation for scalable structured gaussian processes (kiss-gp). In International Conference on Machine Learning, pp 1775–1784. PMLR
Zhang Y, Rădulescu R, Mannion P, Roijers D.M, Nowé A (2020) Opponent modelling for reinforcement learning in multi-objective normal form games. In Proceedings of the 19th International Conference on Autonomous Agents and MultiAgent Systems, pp 2080–2082
Zhang Y, Rădulescu R, Mannion P, Roijers D.M, Nowé A (2020) Opponent modelling using policy reconstruction for multi-objective normal form games. In: Proceedings of the Adaptive and Learning Agents Workshop (ALA-20) at AAMAS
Zintgraf L.M, Roijers D.M, Linders S, Jonker C.M, Nowé A (2018) Ordered preference elicitation strategies for supporting multi-objective decision making. In Proceedings of the 17th International Conference on Autonomous Agents and MultiAgent Systems, pp 1477–1485. International Foundation for Autonomous Agents and Multiagent Systems
Acknowledgements
The authors would like to acknowledge FWO (Fonds Wetenschappelijk Onderzoek) for their support through the SB grants of Timothy Verstraeten (#1S47617N). This research was supported by funding from the Flemish Government under the “Onderzoeksprogramma Artificiële Intelligentie (AI) Vlaanderen” program.
Author information
Authors and Affiliations
Corresponding author
Ethics declarations
Conflict of interest
The authors declare that they have no conflict of interest.
Rights and permissions
About this article
Cite this article
Rădulescu, R., Verstraeten, T., Zhang, Y. et al. Opponent learning awareness and modelling in multi-objective normal form games. Neural Comput & Applic 34, 1759–1781 (2022). https://doi.org/10.1007/s00521-021-06184-3
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s00521-021-06184-3