arXivDaily arXiv每日学术速递 周一至周五更新

期刊&会议

International Joint Conference on Artificial Intelligence · 会议 · Artificial Intelligence

2026-04-23 至 2026-04-23 共收录 1
2604.20074 2026-04-23 cs.LG

Maximum Entropy Semi-Supervised Inverse Reinforcement Learning

最大熵半监督逆强化学习

Julien Audiffren, Michal Valko, Alessandro Lazaric, Mohammad Ghavamzadeh

机构 * CMLA, ENS Cachan(CMLA,ENS巴黎-坎帕尼昂) SequeL team, INRIA Lille - Nord Europe(SequeL团队,INRIA里尔-北欧) Adobe Research(Adobe研究)

AI总结 本文提出MESSI算法,结合最大熵逆强化学习与半监督学习原理,利用未标注数据提升性能,在高速驾驶和网格世界任务中验证了其有效性。

Comments In Proceedings of the 24th International Joint Conference on Artificial Intelligence (IJCAI 2015)

详情

展开后加载摘要…

URL PDF HTML 收藏