1. NCCU Academic Hub
  2. 學術產出
  3. College of Informatics
  4. Proceedings
  5. Item

學術產出-Proceedings

Article View/Open

Publication Export

Google ScholarTM

政大圖書館

Citation Infomation

題名 Heterogeneous AdaBoost with stochastic algorithm selection
作者 Hsu, Kuo-Wei
徐國偉
貢獻者 資訊科學系
關鍵詞 Adaptive boosting; Classification (of information); Information management; Random processes; Stochastic systems; Algorithm selection; Boosting; Ensemble; Ensemble learning; Ensemble learning algorithm; Stochastic algorithms; Training data sets; Learning algorithms
日期 2017-01
上傳時間 3-Aug-2017 14:13:12 (UTC+8)
摘要 We propose an ensemble learning algorithm based on AdaBoost and employing heterogeneous algorithms with a stochastic process for algorithm selection. Diversity is an important factor in ensemble learning and AdaBoost creates diversity by manipulating training data sets. However, we observe that AdaBoost generates training data sets of low diversity in later iterations. Some researchers suggest the employment of heterogeneous algorithms in ensemble learning to achieve better diversity. Following the idea, we extend AdaBoost and propose an algorithm that employs different base learning algorithms in different iterations. The most distinguishing feature of our algorithm is that it selects algorithms using a stochastic process where their earlier performance is considered. The results from experiments on several data sets show the utility of our algorithm: It could outperform AdaBoost on 22 to 33, depending on settings, out of 40 data sets considered in experiments. © 2017 ACM.
關聯 Proceedings of the 11th International Conference on Ubiquitous Information Management and Communication, IMCOM 2017,
11th International Conference on Ubiquitous Information Management and Communication, IMCOM 2017; Beppu; Japan; 5 January 2017 到 7 January 2017; 代碼 126221
資料類型 conference
DOI http://dx.doi.org/10.1145/3022227.3022266
dc.contributor 資訊科學系zh_Tw
dc.creator (作者) Hsu, Kuo-Weien_US
dc.creator (作者) 徐國偉zh_TW
dc.date (日期) 2017-01en_US
dc.date.accessioned 3-Aug-2017 14:13:12 (UTC+8)-
dc.date.available 3-Aug-2017 14:13:12 (UTC+8)-
dc.date.issued (上傳時間) 3-Aug-2017 14:13:12 (UTC+8)-
dc.identifier.uri (URI) http://nccur.lib.nccu.edu.tw/handle/140.119/111624-
dc.description.abstract (摘要) We propose an ensemble learning algorithm based on AdaBoost and employing heterogeneous algorithms with a stochastic process for algorithm selection. Diversity is an important factor in ensemble learning and AdaBoost creates diversity by manipulating training data sets. However, we observe that AdaBoost generates training data sets of low diversity in later iterations. Some researchers suggest the employment of heterogeneous algorithms in ensemble learning to achieve better diversity. Following the idea, we extend AdaBoost and propose an algorithm that employs different base learning algorithms in different iterations. The most distinguishing feature of our algorithm is that it selects algorithms using a stochastic process where their earlier performance is considered. The results from experiments on several data sets show the utility of our algorithm: It could outperform AdaBoost on 22 to 33, depending on settings, out of 40 data sets considered in experiments. © 2017 ACM.en_US
dc.format.extent 347938 bytes-
dc.format.mimetype application/pdf-
dc.relation (關聯) Proceedings of the 11th International Conference on Ubiquitous Information Management and Communication, IMCOM 2017,en_US
dc.relation (關聯) 11th International Conference on Ubiquitous Information Management and Communication, IMCOM 2017; Beppu; Japan; 5 January 2017 到 7 January 2017; 代碼 126221en_US
dc.subject (關鍵詞) Adaptive boosting; Classification (of information); Information management; Random processes; Stochastic systems; Algorithm selection; Boosting; Ensemble; Ensemble learning; Ensemble learning algorithm; Stochastic algorithms; Training data sets; Learning algorithmsen_US
dc.title (題名) Heterogeneous AdaBoost with stochastic algorithm selectionen_US
dc.type (資料類型) conference
dc.identifier.doi (DOI) 10.1145/3022227.3022266
dc.doi.uri (DOI) http://dx.doi.org/10.1145/3022227.3022266