@inproceedings{0f7f2aeacf0547e688cd82c06a596f53,
title = "A linear online guided policy search algorithm",
abstract = "In reinforcement learning (RL), the guided policy search (GPS), a variant of policy search method, can encode the policy directly as well as search for optimal solutions in the policy space. Even though this algorithm is provided with asymptotic local convergence guarantees, it can not work in a online way for conducting tasks in complex environments since it is trained with a batch manner which requires that all of the training samples should be given at the same time. In this paper, we propose an online version for GPS algorithm, which can learn policies incrementally without complete knowledge of initial positions for training. The experiments witness its efficacy on handling sequentially arriving training samples in a peg insertion task.",
keywords = "Online learning, Policy search, Reinforcement learning",
author = "Biao Sun and Fangzhou Xiong and Zhiyong Liu and Xu Yang and Hong Qiao",
note = "Publisher Copyright: {\textcopyright} Springer International Publishing AG 2017.; 24th International Conference on Neural Information Processing, ICONIP 2017 ; Conference date: 14-11-2017 Through 18-11-2017",
year = "2017",
doi = "10.1007/978-3-319-70139-4\_44",
language = "英语",
isbn = "9783319701387",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
publisher = "Springer Verlag",
pages = "434--441",
editor = "Dongbin Zhao and Yuanqing Li and El-Alfy, \{El-Sayed M.\} and Derong Liu and Shengli Xie",
booktitle = "Neural Information Processing - 24th International Conference, ICONIP 2017, Proceedings",
}