Publications-Periodical Articles

Article View/Open

Publication Export

Google ScholarTM

NCCU Library

Citation Infomation

Related Publications in TAIR

題名 A Stochastic Approximate View of Boosting
作者 曹振海;張源俊
日期 2007-09
上傳時間 19-Dec-2008 14:52:05 (UTC+8)
摘要 The boosting as a stochastic approximation algorithm is considered. This new interpretation provides an alternative theoretical framework for investigation. Following the results of stochastic approximation theory a stochastic approximation boosting algorithm, SABoost, is proposed. By adjusting its step sizes, SABoost will have different kinds of properties. Empirically, it is found that SABoost with a small step size will have smaller training and testing errors difference, and when the step size becomes large, it tends to overfit (i.e. bias towards training scenarios). This choice of step size can be viewed as a smooth (early) stopping rule. The performance of AdaBoost is compared and contrasted.
關聯 Computational Statistics & Data Analysis, 52(1), 325-334
資料類型 article
DOI http://dx.doi.org/10.1016/j.csda.2007.06.020
dc.creator (作者) 曹振海;張源俊zh_TW
dc.date (日期) 2007-09en_US
dc.date.accessioned 19-Dec-2008 14:52:05 (UTC+8)-
dc.date.available 19-Dec-2008 14:52:05 (UTC+8)-
dc.date.issued (上傳時間) 19-Dec-2008 14:52:05 (UTC+8)-
dc.identifier.uri (URI) https://nccur.lib.nccu.edu.tw/handle/140.119/18163-
dc.description.abstract (摘要) The boosting as a stochastic approximation algorithm is considered. This new interpretation provides an alternative theoretical framework for investigation. Following the results of stochastic approximation theory a stochastic approximation boosting algorithm, SABoost, is proposed. By adjusting its step sizes, SABoost will have different kinds of properties. Empirically, it is found that SABoost with a small step size will have smaller training and testing errors difference, and when the step size becomes large, it tends to overfit (i.e. bias towards training scenarios). This choice of step size can be viewed as a smooth (early) stopping rule. The performance of AdaBoost is compared and contrasted.-
dc.format application/en_US
dc.language enen_US
dc.language en-USen_US
dc.language.iso en_US-
dc.relation (關聯) Computational Statistics & Data Analysis, 52(1), 325-334en_US
dc.title (題名) A Stochastic Approximate View of Boostingen_US
dc.type (資料類型) articleen
dc.identifier.doi (DOI) 10.1016/j.csda.2007.06.020en_US
dc.doi.uri (DOI) http://dx.doi.org/10.1016/j.csda.2007.06.020en_US