Filters








576 Hits in 7.7 sec

From Few to More: Large-scale Dynamic Multiagent Curriculum Learning [article]

Weixun Wang, Tianpei Yang, Yong Liu, Jianye Hao, Xiaotian Hao, Yujing Hu, Yingfeng Chen, Changjie Fan, Yang Gao
2019 arXiv   pre-print
In this paper, we design a novel Dynamic Multiagent Curriculum Learning (DyMA-CL) to solve large-scale problems by starting from learning on a multiagent scenario with a small size and progressively increasing  ...  However, it is still challenging in large-scale multiagent settings due to the complex dynamics between the environment and agents and the explosion of state-action space.  ...  In this paper, we focus on CL in multiagent RL settings and design a dynamic multiagent curriculum learning to solve large-scale multiagent learning problems.  ... 
arXiv:1909.02790v2 fatcat:fsww4pjwijg6pihi66sphs5fni

From Few to More: Large-Scale Dynamic Multiagent Curriculum Learning

Weixun Wang, Tianpei Yang, Yong Liu, Jianye Hao, Xiaotian Hao, Yujing Hu, Yingfeng Chen, Changjie Fan, Yang Gao
2020 PROCEEDINGS OF THE THIRTIETH AAAI CONFERENCE ON ARTIFICIAL INTELLIGENCE AND THE TWENTY-EIGHTH INNOVATIVE APPLICATIONS OF ARTIFICIAL INTELLIGENCE CONFERENCE  
In this paper, we design a novel Dynamic Multiagent Curriculum Learning (DyMA-CL) to solve large-scale problems by starting from learning on a multiagent scenario with a small size and progressively increasing  ...  However, it is still challenging in large-scale multiagent settings due to the complex dynamics between the environment and agents and the explosion of state-action space.  ...  In this paper, we focus on CL in multiagent RL settings and design a dynamic multiagent curriculum learning to solve large-scale multiagent learning problems.  ... 
doi:10.1609/aaai.v34i05.6221 fatcat:7kwnnbppcrgvbcrmck4mopgviq

Neural MMO: A Massively Multiagent Game Environment for Training and Evaluating Intelligent Agents [article]

Joseph Suarez, Yilun Du, Phillip Isola, Igor Mordatch
2019 arXiv   pre-print
Our environment is well suited to the study of large-scale multiagent interaction: it requires that agents learn robust combat and navigation policies in the presence of large populations attempting to  ...  The emergence of complex life on Earth is often attributed to the arms race that ensued from a huge number of organisms all competing for finite resources.  ...  Thank you to Clare Zhu for substantial contributions to the 3D client code.  ... 
arXiv:1903.00784v1 fatcat:lpl4chdkvffcta5wd3by6almbq

Learning Generalizable Multi-Lane Mixed-Autonomy Behaviors in Single Lane Representations of Traffic [article]

Abdul Rahman Kreidieh, Yibo Zhao, Samyak Parajuli, Alexandre Bayen
2021 arXiv   pre-print
Within this problem, we design a curriculum learning paradigm that exploits the natural extendability of the network to effectively learn behaviors that reduce congestion over long horizons.  ...  Our findings suggest that introducing lane change behaviors that even approximately match trends in more complex systems can significantly improve the generalizability of subsequent learned models to more  ...  For one, to address mismatches that arise from variations in the boundary conditions, we construct a curriculum learning paradigm that scales the performance of policies learned to larger rings, where  ... 
arXiv:2112.04688v2 fatcat:bgvdgngxqfeene7ticl3sky6e4

Cooperative Multi-agent Control Using Deep Reinforcement Learning [chapter]

Jayesh K. Gupta, Maxim Egorov, Mykel Kochenderfer
2017 Lecture Notes in Computer Science  
We also show that recurrent policies, while more difficult to train, outperform feed-forward policies on our evaluation tasks.  ...  Using deep reinforcement learning with a curriculum learning scheme, our approach can solve problems that were previously considered intractable by most multi-agent reinforcement learning algorithms.  ...  The authors would like to thank the anonymous reviewers for their helpful comments.  ... 
doi:10.1007/978-3-319-71682-4_5 fatcat:ie4vvneipjgxbdwngj3bncs6eu

ForMIC: Foraging via Multiagent RL with Implicit Communication [article]

Samuel Shaw, Emerson Wenzel, Alexis Walker, Guillaume Sartoretti
2022 arXiv   pre-print
By relying on clever curriculum learning design, action filtering, and the introduction of non-learning agents to increase the agent density at training time at low computational cost, we develop a minimal  ...  Multi-agent foraging (MAF) involves distributing a team of agents to search an environment and extract resources from it.  ...  A pheromone curriculum may also be looked at as a form of imitation learning, where an agent learns from an expert demonstrator [32] .  ... 
arXiv:2006.08152v4 fatcat:xkmxp4jukvazvm7sytdnaxdfry

A Survey on Transfer Learning for Multiagent Reinforcement Learning Systems

Felipe Leno Da Silva, Anna Helena Reali Costa
2019 The Journal of Artificial Intelligence Research  
For this reason, reusing knowledge that can come from previous experience or other agents is indispensable to scale up multiagent RL algorithms.  ...  However, learning a complex task from scratch is impractical due to the huge sample complexity of RL algorithms.  ...  This work was partially carried out while the first author was affiliated to the Learning Agents Research Group (LARG) at the University of Texas at Austin, TX, USA.  ... 
doi:10.1613/jair.1.11396 fatcat:mn4gw6oh5zgszl6l53fgesei5i

Introducing Electric Power Into a Multidisciplinary Curriculum for Network Industries

M. Ilic, J. Apt, P. Khosla, L.B. Lave, M.G. Morgan, S. Talukdar
2004 IEEE Transactions on Power Systems  
The new curriculum: 1) provides students with a multidisciplinary introduction to the changing problems of the industry; 2) stresses the need for teaching systematic approaches to formulating power system  ...  A qualitatively different graduate level curriculum for teaching electric power systems is needed.  ...  The curriculum will provide them with examples from which they learn to apply the theory to a particular problem.  ... 
doi:10.1109/tpwrs.2003.821014 fatcat:wjyuamue5vg5zlrresu4nuqwfa

The Neural MMO Platform for Massively Multiagent Research [article]

Joseph Suarez, Yilun Du, Clare Zhu, Igor Mordatch, Phillip Isola
2021 arXiv   pre-print
Initial baselines on the platform demonstrate that agents trained in large populations explore more and learn a progression of skills.  ...  We raise other more difficult problems such as many-team cooperation as open research questions which Neural MMO is well-suited to answer.  ...  of learned behaviors and enable even more new directions in multiagent research.  ... 
arXiv:2110.07594v1 fatcat:ev7skquzwvhuxeux46cr7eaej4

Engineering Multi-Agent Systems (Dagstuhl Seminar 12342)

Jürgen Dix, Koen V. Hindriks, Brian Logan, Wayne Wobcke, Marc Herbstritt
2012 Dagstuhl Reports  
As such it was particularly relevant to industrial research. A key objective of the seminar, moreover, has been to establish a roadmap for engineering multiagent systems.  ...  The seminar brought together researchers from both academia and industry to identify the potential for and facilitate convergence towards standards for agent technology.  ...  to the large scale, realistic scenarios found in industry.  ... 
doi:10.4230/dagrep.2.8.74 dblp:journals/dagstuhl-reports/DixHLW12 fatcat:cn2i425enzfohjipslz2aqmen4

Resisting Bureaucracy: A Case Study of Home Schooling

Jean A. Patterson, Ian Gibson, Andrew Koenigs, Michael Maurer, Gladys Ritterhouse, Charles Stockton, Mary Jo Taylor
2007 Journal of Thought  
Curriculum, Educational Materials, and Support Services Publishers of textbooks and large-scale assessments have benefited from the trend toward increasing state and national control over educational decisions  ...  These included dynamic and fluid networks, flexible structures and schedules, responsive pedagogy, tailored curriculum and materials, small classes, and multiage groupings.  ... 
doi:10.2307/jthought.42.3-4.71 fatcat:mptfm23vpjc63jdrfmxmr4pit4

RoboCup as an Introduction to CS Research [chapter]

Peter Stone
2004 Lecture Notes in Computer Science  
This paper proposes using topics central to RoboCup, particularly autonomous agents and multiagent systems, as the subject-matter for a course designed to introduce undergraduate students to all facets  ...  Experiences are presented from the design and implementation of such a course.  ...  Acknowledgements Thanks to Tucker Balch, Jeremy Cooperstock, Gal Kaminka, Manuela Veloso, and José Vidal for ideas from their related courses.  ... 
doi:10.1007/978-3-540-25940-4_25 fatcat:pmsh3bhio5dwrntyngvvyq277m

A Survey on Reinforcement Learning Methods in Character Animation [article]

Ariel Kwiatkowski, Eduardo Alvarado, Vicky Kalogeiton, C. Karen Liu, Julien Pettré, Michiel van de Panne, Marie-Paule Cani
2022 arXiv   pre-print
This paper surveys the modern Deep Reinforcement Learning methods and discusses their possible applications in Character Animation, from skeletal control of a single, physically-based character to navigation  ...  Reinforcement Learning is an area of Machine Learning focused on how agents can be trained to make sequential decisions, and achieve a particular goal within an arbitrary environment.  ...  Acknowledgement This work has received funding from the European Union's Horizon 2020 research and innovation programme under the Marie Skłodowska-Curie grant agreement No 860768 (CLIPE project).  ... 
arXiv:2203.04735v1 fatcat:usnqama2frfwxijpctt6bipivu

A Cooperation Graph Approach for Multiagent Sparse Reward Reinforcement Learning [article]

Qingxu Fu, Tenghai Qiu, Zhiqiang Pu, Jianqiang Yi, Wanmai Yuan
2022 arXiv   pre-print
Next, based on this novel graph structure, we propose a Cooperation Graph Multiagent Reinforcement Learning (CG-MARL) algorithm, which can efficiently deal with the sparse reward problem in multiagent  ...  Multiagent reinforcement learning (MARL) can solve complex cooperative tasks. However, the efficiency of existing MARL methods relies heavily on well-defined reward functions.  ...  We have not discovered general-purpose clusteractions nor the way to design them automatically. We expect that our work can inspire further studies of sparse-reward MARL. VI. ACKNOWLEDGMENTS  ... 
arXiv:2208.03002v1 fatcat:shr6spgsybadxbgi7jjxzwhozq

A Tri-State Study: Is the Middle School Movement Thriving… or Barely Surviving?

John A. Huss, Shannon Eastep
2011 RMLE Online: Research in Middle Level Education  
In sum, the results from this tri-state study suggested that teachers still consider the middle school concept to be quite relevant and applicable.  ...  Random cluster sampling was used to select participants from a population list of districts. One hundred four teachers of 201 (52%) completed the questionnaires.  ...  With many in the middle school movement perceiving the dynamism of the middle school concept to be precarious, at best, additional data are needed in a more accelerated and deliberate fashion to (a) ensure  ... 
doi:10.1080/19404476.2011.11462082 fatcat:absrmu2bgffz3hzkpzrgyslxam
« Previous Showing results 1 — 15 out of 576 results