?url_ver=Z39.88-2004&rft_val_fmt=info%3Aofi%2Ffmt%3Akev%3Amtx%3Adc&rft.title=MAgent%3A+A+many-agent+reinforcement+learning+platform+for+artificial+collective+intelligence&rft.creator=Zheng%2C+L&rft.creator=Yang%2C+J&rft.creator=Cai%2C+H&rft.creator=Zhang%2C+W&rft.creator=Wang%2C+J&rft.creator=Yu%2C+Y&rft.description=We+introduce+MAgent%2C+a+platform+to+support+research+and+development+of+many-agent+reinforcement+learning.+Unlike+previous+research+platforms+on+single+or+multi-agent+reinforcement+learning%2C+MAgent+focuses+on+supporting+the+tasks+and+the+applications+that+require+hundreds+to+millions+of+agents.+Within+the+interactions+among+a+population+of+agents%2C+it+enables+not+only+the+study+of+learning+algorithms+for+agents'+optimal+polices%2C+but+more+importantly%2C+the+observation+and+understanding+of+individual+agent's+behaviors+and+social+phenomena+emerging+from+the+AI+society%2C+including+communication+languages%2C+leaderships%2C+altruism.+MAgent+is+highly+scalable+and+can+host+up+to+one+million+agents+on+a+single+GPU+server.+MAgent+also+provides+flexible+configurations+for+AI+researchers+to+design+their+customized+environments+and+agents.+In+this+demo%2C+we+present+three+environments+designed+on+MAgent+and+show+emerged+collective+intelligence+by+learning+from+scratch.&rft.subject=Reinforcement+learning%3B+multiagent+system%3B+learning+environment%3B&rft.publisher=AAAI&rft.date=2018-02-07&rft.type=Proceedings+paper&rft.language=eng&rft.source=+++++In%3A+++(Proceedings)+32nd+AAAI+Conference+on+Artificial+Intelligence%2C+AAAI+2018.+(pp.+pp.+8222-8223).++AAAI%3A+New+Orleans%2C+LA%2C+USA.+(2018)+++++&rft.format=text&rft.identifier=https%3A%2F%2Fdiscovery.ucl.ac.uk%2Fid%2Feprint%2F10070682%2F1%2F16459-77215-1-PB.pdf&rft.identifier=https%3A%2F%2Fdiscovery.ucl.ac.uk%2Fid%2Feprint%2F10070682%2F&rft.rights=open