Exploiting Reinforcement-Learning for Influence Maximization without Human-Annotated Data

碩士 === 國立臺灣大學 === 資訊網路與多媒體研究所 === 104 === Strategies to choose nodes on a social network to maximize the total influence has been studied for decades. Studies have shown that the greedy algorithm is a competitive strategy and it has been proved to cover at least 63% of the optimal spread. Here we pr...

Full description

Bibliographic Details
Main Authors:	Yen-Hua Huang, 黃彥樺
Other Authors:	Shou-de Lin
Format:	Others
Language:	en_US
Published:	2015
Online Access:	http://ndltd.ncl.edu.tw/handle/24425905249677578391

id	ndltd-TW-104NTU05641002
record_format	oai_dc
spelling	ndltd-TW-104NTU056410022017-06-03T04:41:37Z http://ndltd.ncl.edu.tw/handle/24425905249677578391 Exploiting Reinforcement-Learning for Influence Maximization without Human-Annotated Data 在未標記資料上利用增強式學習解決影響力最大化之問題 Yen-Hua Huang 黃彥樺碩士國立臺灣大學資訊網路與多媒體研究所 104 Strategies to choose nodes on a social network to maximize the total influence has been studied for decades. Studies have shown that the greedy algorithm is a competitive strategy and it has been proved to cover at least 63% of the optimal spread. Here we propose a learning-based framework for influence maximization aiming at outperforming the greedy algorithm in terms of both coverage and efficiency. The proposed reinforcement learning framework combining with a classification model not only alleviates the requirement of the labelled training data, but also allows the influence maximization strategy to be developed gradually and eventually outperforms a basic greedy approach. Shou-de Lin 林守德 2015 學位論文 ; thesis 39 en_US
collection	NDLTD
language	en_US
format	Others
sources	NDLTD
description	碩士 === 國立臺灣大學 === 資訊網路與多媒體研究所 === 104 === Strategies to choose nodes on a social network to maximize the total influence has been studied for decades. Studies have shown that the greedy algorithm is a competitive strategy and it has been proved to cover at least 63% of the optimal spread. Here we propose a learning-based framework for influence maximization aiming at outperforming the greedy algorithm in terms of both coverage and efficiency. The proposed reinforcement learning framework combining with a classification model not only alleviates the requirement of the labelled training data, but also allows the influence maximization strategy to be developed gradually and eventually outperforms a basic greedy approach.
author2	Shou-de Lin
author_facet	Shou-de Lin Yen-Hua Huang 黃彥樺
author	Yen-Hua Huang 黃彥樺
spellingShingle	Yen-Hua Huang 黃彥樺 Exploiting Reinforcement-Learning for Influence Maximization without Human-Annotated Data
author_sort	Yen-Hua Huang
title	Exploiting Reinforcement-Learning for Influence Maximization without Human-Annotated Data
title_short	Exploiting Reinforcement-Learning for Influence Maximization without Human-Annotated Data
title_full	Exploiting Reinforcement-Learning for Influence Maximization without Human-Annotated Data
title_fullStr	Exploiting Reinforcement-Learning for Influence Maximization without Human-Annotated Data
title_full_unstemmed	Exploiting Reinforcement-Learning for Influence Maximization without Human-Annotated Data
title_sort	exploiting reinforcement-learning for influence maximization without human-annotated data
publishDate	2015
url	http://ndltd.ncl.edu.tw/handle/24425905249677578391
work_keys_str_mv	AT yenhuahuang exploitingreinforcementlearningforinfluencemaximizationwithouthumanannotateddata AT huángyànhuà exploitingreinforcementlearningforinfluencemaximizationwithouthumanannotateddata AT yenhuahuang zàiwèibiāojìzīliàoshànglìyòngzēngqiángshìxuéxíjiějuéyǐngxiǎnglìzuìdàhuàzhīwèntí AT huángyànhuà zàiwèibiāojìzīliàoshànglìyòngzēngqiángshìxuéxíjiějuéyǐngxiǎnglìzuìdàhuàzhīwèntí
_version_	1718455078791151616

Exploiting Reinforcement-Learning for Influence Maximization without Human-Annotated Data

Similar Items