174 Hits in 5.5 sec

CM3: Cooperative Multi-goal Multi-stage Multi-agent Reinforcement Learning [article]

Jiachen Yang, Alireza Nakhaei, David Isele, Kikuo Fujimura, Hongyuan Zha
2020 arXiv   pre-print
To address both challenges, we restructure the problem into a novel two-stage curriculum, in which single-agent goal attainment is learned prior to learning multi-agent cooperation, and we derive a new  ...  The complete architecture, called CM3, learns significantly faster than direct adaptations of existing algorithms on three challenging multi-goal multi-agent problems: cooperative navigation in difficult  ...  Multiagent cooperation and competition with deep reinforcement learning. PloS one, 12(4), e0172395. Tan, M. (1993). Multi-agent reinforcement learning: Independent vs. cooperative agents.  ... 
arXiv:1809.05188v3 fatcat:dld5q6ycojcx7elo5ugou26bti

Collaborative Target Search with a Visual Drone Swarm: An Adaptive Curriculum Embedded Multi-stage Reinforcement Learning Approach [article]

Jiaping Xiao, Phumrapee Pisutsin, Mir Feroskhan
2022 arXiv   pre-print
In this work, we propose a data-efficient reinforcement learning-based approach, Adaptive Curriculum Embedded Multi-Stage Learning (ACEMSL), to address the challenges of carrying out a collaborative target  ...  Meanwhile, with multi-stage learning, ACEMSL allows data-efficient training and individual-team reward allocation for the collaborative drone swarm.  ...  In [19] , CM3 (Cooperative Multi-goal, Multi-stage, Multi-agent) was proposed for multi-agent reinforcement learning in 2D space to solve a collaborative navigation problem, where the positions of landmarks  ... 
arXiv:2204.12181v2 fatcat:wo5bis23uzb3zfzzdna2iwf4wq

Local Advantage Actor-Critic for Robust Multi-Agent Deep Reinforcement Learning [article]

Yuchen Xiao, Xueguang Lyu, Christopher Amato
2021 arXiv   pre-print
Policy gradient methods have become popular in multi-agent reinforcement learning, but they suffer from high variance due to the presence of environmental stochasticity and exploring agents (i.e., non-stationarity  ...  To this end, we propose a new multi-agent policy gradient method, called Robust Local Advantage (ROLA) Actor-Critic.  ...  Zha, “Cm3: Proceedings of the Conference on Robot Learning, November 2020. Cooperative multi-goal multi-stage multi-agent reinforcement learn- [2] Y. C.  ... 
arXiv:2110.08642v3 fatcat:cgkvfeqzdjd6znwobewgarqct4

Interaction-aware Decision Making with Adaptive Strategies under Merging Scenarios [article]

Yeping Hu, Alireza Nakhaei, Masayoshi Tomizuka, Kikuo Fujimura
2020 arXiv   pre-print
A single policy is learned under the multi-agent reinforcement learning (MARL) setting via the curriculum learning strategy, which enables the agent to automatically infer other drivers' various behaviors  ...  A masking mechanism is also proposed to prevent the agent from exploring states that violate common sense of human judgment and increase the learning efficiency.  ...  [12] proposed a cooperative multi-goal multi-stage multiagent reinforcement learning (CM3) approach, which learns an actor and a double critic shared by all agents.  ... 
arXiv:1904.06025v2 fatcat:kvjwwdmwk5blzbypv434mj6qgy

A Review of Cooperative Multi-Agent Deep Reinforcement Learning [article]

Afshin OroojlooyJadid, Davood Hajinezhad
2021 arXiv   pre-print
Deep Reinforcement Learning has made significant progress in multi-agent systems in recent years.  ...  In particular, we have focused on five common approaches on modeling and solving cooperative multi-agent reinforcement learning problems: (I) independent learners, (II) fully observable critic, (III) value  ...  A graph convolutional reinforcement learning for cooperative multi-agent is proposed.  ... 
arXiv:1908.03963v4 fatcat:s2umqzxmqrhntkev3f6k554cv4

Curriculum Learning for Reinforcement Learning Domains: A Framework and Survey [article]

Sanmit Narvekar and Bei Peng and Matteo Leonetti and Jivko Sinapov and Matthew E. Taylor and Peter Stone
2020 arXiv   pre-print
Reinforcement learning (RL) is a popular paradigm for addressing sequential decision tasks in which the agent has only limited environmental feedback.  ...  In this article, we present a framework for curriculum learning (CL) in reinforcement learning, and use it to survey and classify existing CL methods in terms of their assumptions, capabilities, and goals  ...  Part of this work has taken place in the Learning Agents Research Group (LARG) at the Artificial Intelligence Laboratory, The University of Texas at Austin. LARG re-  ... 
arXiv:2003.04960v2 fatcat:iacmqeb7jjeezpo27jsnzuqb7u

'Value for Whom, by Whom': Investigating Value Constructs in Non-Profit Project Portfolios

Karyne Cheng Siew Ang, Shankar Sankaran, Catherine Patricia Killen
2016 Project Management Research and Practice  
The study extends our knowledge about strategic value and multi-stakeholder management in the non-profit sector.  ...  Next, it deliberates the extent to which multi-stakeholder perspectives of value are discussed in the literature.  ...  (CM3) You learn a lot by being involved in things. You actually grow as a person. (CM4) A further value construct involved the direct personal engagement of members with the project.  ... 
doi:10.5130/pmrp.v3i0.5038 fatcat:57p5kgysdbffdby2erdrtiehqm

Defense Advanced Research Projects Agency (Darpa) Fiscal Year 2019 Budget Estimates

Department Of Defense Comptroller's Office
2018 Zenodo  
As a result, autonomous systems enabled by machine learning (e.g., deep neural nets for perception, reinforcement learning for control policies, and online model learning) lack rigorous safety assurance  ...  -Explore formal approaches to verify correctness properties of autonomous software agents and use machine learning or similar artificial intelligence techniques to ensure safe and reliable autonomous agent  ... 
doi:10.5281/zenodo.1215599 fatcat:kxduvg6a5nflhf6iilkwnolobu

ECR 2012 Book of Abstracts - A - Postergraduate Educational Programme

2012 Insights into Imaging  
To learn how to differentiate ischaemia from inflammation.  ...  Implementation and evaluation of multimodality molecular imaging technology producing multiparametric and multivariable techniques is reinforced.  ...  To learn about the goal of each medical specialty in pancreatic tumours. A-024 16:05 What the surgeon needs to know C.  ... 
doi:10.1007/s13244-012-0153-4 pmid:22696127 pmcid:PMC3481066 fatcat:te6ctbtakzh5njsw43geghw3ta


Cristina Paca, Carmen Fenollosa, Raymond Gemen, Barbaros Corekoglu, Jacqueline Broerse
2020 Zenodo  
Step B: Formulating learning goals. As a subsequent step, learning goals can be formulated for the competences identified.  ...  None; learning goals and competences can be achieved during the course of the project OPTIONS FOR MULTI-STAKEHOLDER ENGAGEMENT Stakeholder Role envisaged in the activity There is room for all types  ...  Tell the people that in this workshop all of the participants will make their own vacuum pump, test it out and learn how it connects with food preservation.  ... 
doi:10.5281/zenodo.4601638 fatcat:rvhk2o7gyvd63cq2vbkzl6iawe


2020 Zenodo  
Multi-stage sampling procedure was used to select the sample for the study.  ...  Multi-stage sampling procedure was used to select the sample for the study.  ...  Forgetting is characteristic for this stage, and a systematic study of the person with the goal is appropriate.  ... 
doi:10.5281/zenodo.4031845 fatcat:qedklhrea5fnfoknixls3yslqm

Advancing Cutting Technology

G. Byrne, D. Dornfeld, B. Denkena
2003 CIRP annals  
Material removal processes can take place at considerably higher performance levels in the range up to Qw = 150 -1500 cm3/min for most workpiece materials at cutting speeds up to some 8.000 m/min.  ...  This remains an illusive goal due to many challenges.  ...  Now it is possible by CVD and PCVD techniques to synthesise new multi-component and multi-phase coatings for cemented carbide inserts and tools.  ... 
doi:10.1016/s0007-8506(07)60200-5 fatcat:mtvdom63drfxjjlh46umhkfnky

Integrating Climate Change into the Environmental Assessment Process: What is the Situation in African Francophone Countries?

Tchindjang Mesmin
2018 Environment and Ecology Research  
Three global climate models used are ECHAM5 / MPIOM, CNRM-CM3 and IPSL-CM4.  ...  The next example is provided by German technical cooperation tool.  ... 
doi:10.13189/eer.2018.060302 fatcat:l7f2mjbkw5bcncibqtrm7dstx4

Training Manual on Value Chain Analysis of Dryland Agricultural Commodities

Amarender A. Reddy
2013 Social Science Research Network  
Features Producer Cooperative Producer Company Registration Cooperative Societies Act Companies Act Objectives Single Object Multi Object Membership Open only to individuals and cooperatives  ...  Cooperatives: Cooperatives are the structures owned and managed by the producers.  ... 
doi:10.2139/ssrn.2281677 fatcat:jptwz3yegjexpjfgi6oo3zfmsm

Monitoring the Application [chapter]

Adam Shackelford
2015 Beginning Amazon Web Services with Node.js  
We assumed a mean density of the ocean to be 1.0 g/cm3, a mean density of the crust to be 2.75 g/cm3, and a mean density of the mantle to be 3.3 g/cm3.  ...  Most recently, we have extended Pursuit Evasion Game into a multi-agent pursuit Evasion Game in which multiple robotic pursuers collectively determine the location of multiple evaders, and try to corral  ...  survey of the programmatic features of research experiences for undergraduates with the intent of learning what promotes women!  ... 
doi:10.1007/978-1-4842-0653-9_7 fatcat:b6s3wv3jcvf6xiniouokdhwtba
« Previous Showing results 1 — 15 out of 174 results