Learning to Plan Hierarchically from Curriculum

IEEE Robotics and Automation Letters(2019)

引用 6|浏览0
暂无评分
摘要
We present a framework for learning to plan hierarchically in domains with unknown dynamics. We enhance planning performance by exploiting problem structure in several ways: First, we simplify the search over plans by leveraging knowledge of skill objectives; second, shorter plans are generated by enforcing aggressively hierarchical planning; and third, we learn transition dynamics with sparse loc...
更多
查看译文
关键词
Planning,Robots,Cognition,Heuristic algorithms,Trajectory,Task analysis,Markov processes
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络
Chat Paper
正在生成论文摘要