Filters








26,320 Hits in 8.0 sec

Strategically Efficient Exploration in Competitive Multi-agent Reinforcement Learning [article]

Robert Loftin, Aadirupa Saha, Sam Devlin, Katja Hofmann
2021 arXiv   pre-print
High sample complexity remains a barrier to the application of reinforcement learning (RL), particularly in multi-agent systems.  ...  To address this issue, we introduce a formal notion of strategically efficient exploration in Markov games, and use this to develop two strategically efficient learning algorithms for finite Markov games  ...  * This work done while at Microsoft Research Cambridge In this work, we focus on efficient exploration for reinforcement learning in competitive multi-agent settings.  ... 
arXiv:2107.14698v1 fatcat:lpbjs45eevcpvodibns7kthkym

A Survey of Deep Reinforcement Learning in Video Games [article]

Kun Shao, Zhentao Tang, Yuanheng Zhu, Nannan Li, Dongbin Zhao
2019 arXiv   pre-print
Therefore, we also discuss some key points when applying DRL methods to this field, including exploration-exploitation, sample efficiency, generalization and transfer, multi-agent learning, imperfect information  ...  Deep reinforcement learning (DRL) has made great achievements since proposed.  ...  Reactor [39] is a sample-efficient and numerical efficient reinforcement learning agent based on a multi-step return off-policy actor-critic architecture.  ... 
arXiv:1912.10944v2 fatcat:fsuzp2sjrfcgfkyclrsyzflax4

Comparison of Different Auction Pricing Rules in the Electricity Market

Ly Fie Sugianto, Kevin Zhigang Liao
2014 Modern Applied Science  
The Simulated-Annealing Q-learning algorithm has been adopted as the learning mechanism for the agents so they can maximize their profit using strategic bidding.  ...  The objective of this study is to investigate the impact of different auction pricing rules on the market performance in the context of the competitive electricity market.  ...  electricity price which can indirectly enhance the competitiveness of other commodities in many other industries in the region.  ... 
doi:10.5539/mas.v8n1p147 fatcat:noquaae4fnhhzanktatusyhedy

Autocurricula and the Emergence of Innovation from Social Interaction: A Manifesto for Multi-Agent Intelligence Research [article]

Joel Z. Leibo, Edward Hughes, Marc Lanctot, Thore Graepel
2019 arXiv   pre-print
Here we explore the hypothesis that multi-agent systems sometimes display intrinsic dynamics arising from competition and cooperation that provide a naturally emergent curriculum, which we term an autocurriculum  ...  Evolution has produced a multi-scale mosaic of interacting adaptive units.  ...  In addition, the first author would like to thank all the speakers, organizers, and attendees of the 2014 "Are there limits to evolution?"  ... 
arXiv:1903.00742v2 fatcat:xbx5ybhxkbhn7evdguxobyd2yi

Modeling the Strategic Bidding of the Producers in Competitive Electricity Markets with the Watkins' Q (lambda) Reinforcement Learning

Yuchao Ma, Ettore F. Bompard, Roberto Napoli, Jiang Chuanwen
2006 International Journal of Emerging Electric Power Systems  
The decision making process of the producers and their interactions in the market are a typical complex problem that is difficult to model explicitly, and can be studied with a multi agents approach.  ...  Competition has been introduced in the last decade into the electricity markets and is presently underway in many countries.  ...  Compared with the single agent learning case the multi agents learning process is a complexity framework in which the "environment" is affected by all the agents' strategic behaviors.  ... 
doi:10.2202/1553-779x.1258 fatcat:ybilses2drfv3iww73uzyg5dpe

Evaluation of Market Rules Using a Multi-Agent System Method

Nan-Peng Yu, Chen-Ching Liu, J. Price
2010 IEEE Transactions on Power Systems  
Index Terms-Electricity market, market design, market power mitigation, multi-agent system, Q-learning.  ...  Simulation results show that in the absence of market power mitigation, generation company (GENCO) agents facilitated by Q-learning are able to exploit the market flaws and make significantly higher profits  ...  Sheblé's for the contribution in the initial stage of this research.  ... 
doi:10.1109/tpwrs.2009.2030379 fatcat:gozr2aibjvb5valeaghw7btakq

Reinforcement Learning for Argumentation: Describing a PhD Research

Sultan Alahmari, Tommy Yuan, Daniel Kudenko
2017 International Conference on Artificial Intelligence and Law  
In multi-agent systems, each agent interacts with the environment and communicates with other agents in order to achieve the designated goal.  ...  Efficiency is related to whether the agent can learn within a limited or insufficient time. So the aim is to find out if the agent can learn rapidly or not.  ... 
dblp:conf/icail/AlahmariYK17 fatcat:axorb2fsqjcsbednd62qom5hc4

Strategic Maneuver and Disruption with Reinforcement Learning Approaches for Multi-Agent Coordination [article]

Derrik E. Asher, Anjon Basak, Rolando Fernandez, Piyush K. Sharma, Erin G. Zaroukian, Christopher D. Hsu, Michael R. Dorothy, Thomas Mahre, Gerardo Galindo, Luke Frerichs, John Rogers, John Fossaceca
2022 arXiv   pre-print
Reinforcement learning (RL) approaches can illuminate emergent behaviors that facilitate coordination across teams of agents as part of a multi-agent system (MAS), which can provide windows of opportunity  ...  Therefore, as part of a defense strategy, friendly forces must use strategic maneuvers and disruption to gain superiority in complex multi-faceted domains such as multi-domain operations (MDO).  ...  MAS in MDO.  ... 
arXiv:2203.09565v1 fatcat:4getezz76ve5nf32fj2izn7bem

Strategic Interaction Multi-Agent Deep Reinforcement Learning

Wenhong Zhou, Jie Li, Yiting Chen, Lincheng Shen
2020 IEEE Access  
INDEX TERMS Multi-agent deep reinforcement learning, scalability, local interaction, large scale.  ...  Despite the proliferation of multi-agent deep reinforcement learning (MADRL), most existing typical methods do not scale well to the dynamics of agent populations.  ...  INTRODUCTION Multi-agent reinforcement learning (MARL) sophisticatedly combines the game theory, multi-agent system and reinforcement learning (RL).  ... 
doi:10.1109/access.2020.3005734 fatcat:4f7qfuidbzfojjrxm2clgu7wci

Page 3927 of Mathematical Reviews Vol. , Issue 2003e [page]

2003 Mathematical Reviews  
The authors explore the strategic trade-offs that this commitment ability and the multiplicity of tasks provide.  ...  Each player attempts to improve her/his average score by adjusting the frequency of the three possible responses, using reinforcement learning.  ... 

Recursive Reasoning Graph for Multi-Agent Reinforcement Learning

Xiaobai Ma, David Isele, Jayesh K. Gupta, Kikuo Fujimura, Mykel J. Kochenderfer
2022 PROCEEDINGS OF THE THIRTIETH AAAI CONFERENCE ON ARTIFICIAL INTELLIGENCE AND THE TWENTY-EIGHTH INNOVATIVE APPLICATIONS OF ARTIFICIAL INTELLIGENCE CONFERENCE  
Multi-agent reinforcement learning (MARL) provides an efficient way for simultaneously learning policies for multiple agents interacting with each other.  ...  This paper adopts a recursive reasoning model in a centralized-training-decentralized-execution framework to help learning agents better cooperate with or compete against others.  ...  Multi-Agent Actor-Critic The actor-critic framework is a common training structure in single-agent reinforcement learning.  ... 
doi:10.1609/aaai.v36i7.20733 fatcat:fnmnzboidzbbplpjclc3x4wsaq

Cooperative and Competitive Biases for Multi-Agent Reinforcement Learning [article]

Heechang Ryu, Hayong Shin, Jinkyoo Park
2021 arXiv   pre-print
Training a multi-agent reinforcement learning (MARL) algorithm is more challenging than training a single-agent reinforcement learning algorithm, because the result of a multi-agent task strongly depends  ...  For a cooperative and competitive environment, there are generally two groups of agents: cooperative-agents and competitive-agents.  ...  Recently, to enhance the performance of multi-agent reinforcement learning (MARL) algorithms, which are extensions of the RL algorithms to multi-agent settings, various strategies have been proposed.  ... 
arXiv:2101.06890v1 fatcat:a4cpts7rgjaffjxnli3igacizy

State-of-the-Art and Open Challenges in RTS Game-AI and Starcraft

Khan Adil, Feng Jiang, Shaohui Liu, Worku Jifara, Zhihong Tian, Yunsheng Fu
2017 International Journal of Advanced Computer Science and Applications  
In such games, each player needs to utilize resources efficiently, which includes managing different types of soldiers, units, equipment's, economic status, positions and the uncertainty during the combat  ...  Finally, we conclude by emphasizing on game 'CIG & AIIDE' competitions along with open research problems and questions in the context of RTS Game-AI, where some of the problems and challenges are mostly  ...  In the proposed method "Q-learning" and "SARSA" algorithms are used with generalized reward functions to train the reinforcement learning agent.  ... 
doi:10.14569/ijacsa.2017.081203 fatcat:3i5ebofonzeidktjbjm6ptbora

Using Multi-Agent Reinforcement Learning in Auction Simulations [article]

Medet Kanmaz, Elif Surer
2020 arXiv   pre-print
The results show that using a multi-agent reinforcement learning strategy improves the outcomes of the auction simulations.  ...  In this study, the strategic (rational) agents created by reinforcement learning algorithms are supposed to be bidder agents in various types of auction mechanisms such as British Auction, Sealed Bid Auction  ...  In this paper, multi-agent reinforcement learning agents in different auction setups were created.  ... 
arXiv:2004.02764v1 fatcat:haxdupddxbg4dazwrv5qbq4xxe

eSense 2.0: Modeling Multi-agent Biomimetic Predation with Multi-layered Reinforcement Learning [chapter]

D. Michael Franklin, Derek Martin
2019 Advances in Biochemical Engineering/Biotechnology  
These independent layers split the learning objectives across multiple layers, avoiding the learning-confusion common in many multi-agent systems.  ...  yet simplistic reinforcement learning algorithm that employs model-based behavior across multiple learning layers.  ...  This is especially true in multi-agent systems where the desired interactions take on a strategic, intelligent meaning.  ... 
doi:10.1007/978-3-030-12385-7_35 fatcat:iglr4bcssjhapbiirlrwusdbiu
« Previous Showing results 1 — 15 out of 26,320 results