Metareview: As per R3: This paper presents a novel approach for doing hierarchical deep RL (HRL) on UMDPs by: (a) use of hindsight experience replay at ...
確定! 回上一頁