應用粒子群最佳化分群演算法於遺漏值問題之研究

NHUIR > College of Science and Technology > Department of Information Management > Disserations and Theses > Item 987654321/19550

Please use this identifier to cite or link to this item: http://nhuir.nhu.edu.tw/handle/987654321/19550

Title:	應用粒子群最佳化分群演算法於遺漏值問題之研究
Other Titles:	The Study of Employing PSO Clustering into the Problem of Missing Values
Authors:	林宜德 Lin, Yi-te
Contributors:	資訊管理學系碩士班邱宏彬 Hung-pin Chiu
Keywords:	群集分析;粒子群最佳化;遺漏值 particle swarm optimization;missing value;clustering analysis
Date:	2010
Issue Date:	2015-03-16 11:42:34 (UTC+8)
Abstract:	在資料採礦的過程中，原始來源資料中的資料遺漏或缺失會造成分析結果異常與偏誤，若是用於企業組織決策支援，可能造成組織決策錯誤，影響企業經營績效。本研究以群集分析技術為基礎，選擇粒子群最佳化演算法發展一套遺漏值推估模組，並提出三個改良方法，透過實驗探討其改良成效。另外針對資料筆數較多之較大型資料集，本研究提出一個兩階段分群推估法，經過實驗證實可有效提昇演算效能。　　In the data mining process, the missing data of the original source will cause analysis results excepted and errors. If the decision support for the organization may lead to organizational decision-making errors, and affect business performance. In this study, cluster analysis based on particle swarm optimization choose to develop a set of missing value estimation module, and propose three improved methods , improvement through the experimental results. For large data sets, this study proposes a two-stage cluster estimation method, after a test proved to be effective to enhance routing performance.
Appears in Collections:	[資訊管理學系] 博碩士論文

Files in This Item:

File	Description	Size	Format
098NHU05396032-001.pdf		2088Kb	Adobe PDF	32	View/Open
index.html		0Kb	HTML	259	View/Open

Loading...