以鑑別分析測量分布間之接近度; Measure the Closeness of Density Functions by Discriminant Analysis

Please use this identifier to cite or link to this item: http://ir.lib.ncu.edu.tw/handle/987654321/7866

Title:	以鑑別分析測量分布間之接近度;Measure the Closeness of Density Functions by Discriminant Analysis
Authors:	鄭振成;Chen-Cheng Cheng
Contributors:	數學研究所
Keywords:	鑑別分析;Discriminant Analysis
Date:	2006-12-29
Issue Date:	2009-09-22 11:07:36 (UTC+8)
Publisher:	國立中央大學圖書館
Abstract:	本文討論雙指數分布，柯西分布，分布(其中表自由度為4之分布)和常態分布之間的接近程度，我們以鑑別分析作為判斷的標準，基本概念如下：較接近常態分布之分布在作常態性鑑別分析時鑑別正確率應較接近常態分布之鑑別正確率。較接近雙指數分布之分布在作雙指數性鑑別分析時鑑別正確率應較接近雙指數分布之鑑別正確率。較接近柯西分布之分布在作柯西性鑑別分析時鑑別正確率應較接近柯西分布之鑑別正確率。我們將用計算機模擬方式比較鑑別正確率，藉以判斷分布間之接近程度。 In this paper , we discuss the closeness of two density functions by discriminant analysis . The idea is as follows . Consider the discriminant analysis based on f , for those g that are close to f , the error rate of discriminantions for g will be close to error rate of discriminantions for f . We will make comparisons between transformed Normal , Double Exponential , Cauchy and t distributions by simulations .
Appears in Collections:	[Graduate Institute of Mathematics] Electronic Thesis & Dissertation

Files in This Item:

File	Size	Format

社群 sharing