東華大學圖書館 |

語系: 繁體中文

說明(常見問題)

回圖書館首頁

手機版館藏查詢

登入

回首頁

切換: 標籤 | MARC模式 | ISBD

FindBook

Google Book

Amazon

博客來

Reinforcement Learning and Relational Learning with Applicationsin Mobile-health and Knowledge Graph.

紀錄類型:	書目-電子資源 : Monograph/item
正題名/作者:	Reinforcement Learning and Relational Learning with Applicationsin Mobile-health and Knowledge Graph./
作者:	Zhang, Sheng.
出版者:	Ann Arbor : ProQuest Dissertations & Theses, : 2021,
面頁冊數:	87 p.
附註:	Source: Dissertations Abstracts International, Volume: 83-05, Section: B.
Contained By:	Dissertations Abstracts International83-05B.
標題:	Computer & video games. -
電子資源:	http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=28747730
ISBN:	9798494448491

Reinforcement Learning and Relational Learning with Applicationsin Mobile-health and Knowledge Graph.
Zhang, Sheng.

Reinforcement Learning and Relational Learning with Applicationsin Mobile-health and Knowledge Graph. - Ann Arbor : ProQuest Dissertations & Theses, 2021 - 87 p.

Source: Dissertations Abstracts International, Volume: 83-05, Section: B.

Thesis (Ph.D.)--North Carolina State University, 2021.

This item must not be sold to any third party vendors.

Reinforcement learning is a general technique that allows an agent to learn the policy to interact with an environment. The goodness of a policy is measured by its value function starting from some initial state. In this thesis, we first construct confidence intervals (CIs) for a policy's value in infinite horizon settings where the number of decision points diverges to infinity. We propose to model the action-value state function (Q-function) associated with a policy based on series/sieve method to derive its confidence interval. When the target policy depends on the observed data as well, we propose a SequentiAl Value Evaluation (SAVE) method to recursively update the estimated policy and its value estimator.To extend the application of reinforcement learning in logical world, we then propose a knowledgeguided reinforcement learning framework for open attribute value extraction. Informed by relevant knowledge in KG, we trained a deep Q-network to sequentially compare extracted answers to improve extraction accuracy. The proposed framework is applicable to different information extraction system.Lastly, we study the underlying structure of the large-scale graph as relational learning. Specifically, we consider networks with "grouped communities" (or "the groups structure"), where nodes within grouped communities are densely connected and nodes across grouped communities are relatively loosely connected, while nodes belonging to the same group but different communities can be either densely or loosely connected. We incorporate the group structure in the stochastic blockmodel and propose a novel divide-and-conquer algorithm to detect the community structure. We show that the proposed method can recover both the group structure and the community structure asymptotically. Numerical studies demonstrate that the proposed method can reduce the computational cost significantly while still achieving competitive performance.

ISBN: 9798494448491Subjects--Topical Terms:

3548317
Computer & video games.

Reinforcement Learning and Relational Learning with Applicationsin Mobile-health and Knowledge Graph.
LDR:03115nmm a2200361 4500 001 2344680
005 20220531064624.5
008 241004s2021 ||||||||||||||||| ||eng d
020 $a 9798494448491
035 $a (MiAaPQ)AAI28747730
035 $a (MiAaPQ)NCState_Univ18402039000
035 $a AAI28747730
040 $a MiAaPQ $c MiAaPQ
100 1 $a Zhang, Sheng. $3 1019185
245 1 0 $a Reinforcement Learning and Relational Learning with Applicationsin Mobile-health and Knowledge Graph.
260 1 $a Ann Arbor : $b ProQuest Dissertations & Theses, $c 2021
300 $a 87 p.
500 $a Source: Dissertations Abstracts International, Volume: 83-05, Section: B.
500 $a Advisor: Chi, Min;Chi, Eric;Lu, Wenbin;Song, Rui.
502 $a Thesis (Ph.D.)--North Carolina State University, 2021.
506 $a This item must not be sold to any third party vendors.
520 $a Reinforcement learning is a general technique that allows an agent to learn the policy to interact with an environment. The goodness of a policy is measured by its value function starting from some initial state. In this thesis, we first construct confidence intervals (CIs) for a policy's value in infinite horizon settings where the number of decision points diverges to infinity. We propose to model the action-value state function (Q-function) associated with a policy based on series/sieve method to derive its confidence interval. When the target policy depends on the observed data as well, we propose a SequentiAl Value Evaluation (SAVE) method to recursively update the estimated policy and its value estimator.To extend the application of reinforcement learning in logical world, we then propose a knowledgeguided reinforcement learning framework for open attribute value extraction. Informed by relevant knowledge in KG, we trained a deep Q-network to sequentially compare extracted answers to improve extraction accuracy. The proposed framework is applicable to different information extraction system.Lastly, we study the underlying structure of the large-scale graph as relational learning. Specifically, we consider networks with "grouped communities" (or "the groups structure"), where nodes within grouped communities are densely connected and nodes across grouped communities are relatively loosely connected, while nodes belonging to the same group but different communities can be either densely or loosely connected. We incorporate the group structure in the stochastic blockmodel and propose a novel divide-and-conquer algorithm to detect the community structure. We show that the proposed method can recover both the group structure and the community structure asymptotically. Numerical studies demonstrate that the proposed method can reduce the computational cost significantly while still achieving competitive performance.
590 $a School code: 0155.
650 4 $a Computer & video games. $3 3548317
650 4 $a Airlines. $3 743401
650 4 $a Teaching methods. $3 595505
650 4 $a Algorithms. $3 536374
650 4 $a Confidence intervals. $3 566017
650 4 $a Telemedicine. $3 841772
650 4 $a Decision making. $3 517204
650 4 $a Car pools. $3 3683472
650 4 $a Robotics. $3 519753
650 4 $a Medical research. $2 bicssc $3 1556686
650 4 $a Computer science. $3 523869
650 4 $a Education. $3 516579
650 4 $a Medicine. $3 641104
650 4 $a Pedagogy. $3 2122828
650 4 $a Recreation. $3 535376
690 $a 0771
690 $a 0984
690 $a 0515
690 $a 0564
690 $a 0456
690 $a 0814
710 2 $a North Carolina State University. $3 1018772
773 0 $t Dissertations Abstracts International $g 83-05B.
790 $a 0155
791 $a Ph.D.
792 $a 2021
793 $a English
856 4 0 $u http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=28747730