Hsiuping University of Science and Technology Institutional Repository : Item 310993100/3851

English | 正體中文 | 简体中文 | Items with full text/Total items : 4342/7641
Visitors : 3555479 Online Users : 145

RC Version 3.2 © Powered By DSPACE, MIT. Enhanced by NTU Library IR team.

Scope Adv. Search

Login ‧ Upload ‧ Help ‧ About ‧ Administer

HUSTIR > College of Engineering > Department of Electrical Engineering & Graduate Institute > Theses and Dissertations > Item 310993100/3851

Please use this identifier to cite or link to this item: http://ir.hust.edu.tw/dspace/handle/310993100/3851

Title:	自我分裂型K-平均模糊類神經系統支持向量回歸用於照度估計 Self-Splitting K-Means Generated Neural Fuzzy System with Support Vector Regression for Illumination Estimation
Authors:	莊易蒼
Contributors:	電機工程研究所
Keywords:	自我分裂型支持向量回歸演算法
Date:	2012-06
Issue Date:	2013-11-19T06:30:48Z
Abstract:	本論文提出一個使用自我分裂型的K–Means演算法與支持向量回歸（SSKNFS-SVR）結合的類神經模糊系统的方法來估計日照流明數。SSKNFS-SVR後鑑部為單值或TSK函數型。對於結構學習，自我分裂型的K–Means演算法是將輸入的訓練數據分群，並決定規則數。對於參數學習，採用線性支持向量回歸（SVR）演算法，調整後鑑部的自由參數。使用SVR支持向量回歸作參數學習的動機是提高SSKNFS- SVR的廣義能力。提出本論文的目的是將SSKNFS- SVR自我分裂型群聚模糊類神經支持向量回歸估算在不同氣候環境下教室的日照模型。 This paper proposes a Self-Splitting K-Means generated Neural Fuzzy System with Support Vector Regression (SSKNFS-SVR). The consequent layer in SSKNFS-SVR is a Takagi-Sugeno-Kang (TSK)-type or singleton consequent. For structure learning, a Self-Splitting K-Means algorithm clusters the input training data and determines the rule number. For parameter learning, a linear support vector regression (SVR) algorithm is proposed to tune free parameters in the consequent part. The motivation for using SVR for parameter learning is to improve the SSKNFS-SVR generalization ability. This paper demonstrates the capabilities of SSKNFS-SVR by conducting simulations in illumination estimation approximations.
Description:	指導教授:謝承達
Appears in Collections:	[Department of Electrical Engineering & Graduate Institute] Theses and Dissertations

Files in This Item:

File	Size	Format
index.html	0Kb	HTML	288	View/Open

Loading...