Randomized Linear Programming Solves the Markov Decision Problem in Nearly Linear (Sometimes Sublinear) Time
Randomized Linear Programming Solves the Markov Decision Problem in Nearly Linear (Sometimes Sublinear) Time
复制标题
随机线性规划在近线性(有时是次线性)时间内解决马尔可夫决策问题
DOI:
10.1287/moor.2019.1000
复制
发表时间:
2020
期刊:
影响因子:
--
通讯作者:
Mengdi Wang
中科院分区:
文献类型:
--
作者:
Mengdi Wang
We propose a novel randomized linear programming algorithm for approximating the optimal policy of the discounted-reward and average-reward Markov decision problems. By leveraging the value–policy ...