國立高雄大學圖資館 |

Language: English

Back

PLS2演算法之比較 = A Comparison of PLS2 Al...

國立高雄大學統計學研究所

PLS2演算法之比較 = A Comparison of PLS2 Algorithms

Record Type:	Language materials, printed : monographic
Paralel Title:	A Comparison of PLS2 Algorithms
Author:	韓佩君,
Secondary Intellectual Responsibility:	國立高雄大學
Place of Publication:	[高雄市]
Published:	撰者;
Year of Publication:	[民99]2010
Description:	64面圖，表 : 30公分;
Subject:	偏最小平方法
Subject:	partial least squares
Online resource:	http://handle.ncl.edu.tw/11296/ndltd/76968173275864377896
Summary:	偏最小平方法(PLS)為多元線性迴歸、主成分分析與典型相關分析的綜合體，它可以處理小樣本以及存在共線性的資料。本論文主要比較近年來PLS2 演算法則之不同點。首先，我們介紹在本文中即將討論的幾種PLS2 與SIMPLS 的演算過程，並使用兩筆資料作實例演算。從實例演算的結果中可知，權重向量單位化與否對迴歸係數的估計會造成結果不同。我們修訂演算法的某些步驟，使它們合乎理論上的推導，最後，修訂之後的PLS 演算法再次使用同樣兩筆資料來進行比較，並以累積解釋變異來選取潛在變數之個數，以迴歸係數透過潛在變數解釋獨立變數與因變數的關係。 Partial least squares(PLS) is the complex of the multiple linear regression, principal component analysis and canonical correlation analysis. It can deal with small sample and the collinearity data. This paper is mainly the comparison of the PLS2 algorithm. First, we introduce several PLS2 process rules and SIMPLS discussed in this article, and use the two data for instance. By the examples, whether making a unit of weight vector or not will cause different to results of regression coefficient of different algorithms. We revised the algorithms to make them to be consistent with the theoretical derivation. After the revision of the PLS algorithm is used again to carry out the same two data sets in comparison. The cumulative explained variance is to select the number of factors, and the regression coefficients are to explain the relationship between independent variables and dependent variables through latent variables.

PLS2演算法之比較 = A Comparison of PLS2 Algorithms
韓, 佩君

PLS2演算法之比較 = A Comparison of PLS2 Algorithms / 韓佩君撰 - [高雄市] : 撰者, [民99]2010. - 64面 ; 圖，表 ; 30公分.
參考書目：面.
偏最小平方法partial least squares

PLS2演算法之比較 = A Comparison of PLS2 Algorithms
LDR:02801nam0a2200277 450 001 272932
005 20170214090245.0
009 272932
010 0 $b 精裝
010 0 $b 平裝
100 $a 20170214y2010 k y0chiy05 e
101 0 $a chi $d chi $d eng
102 $a tw
105 $a ak am 000yy
200 1 $a PLS2演算法之比較 $d A Comparison of PLS2 Algorithms $f 韓佩君撰
210 $a [高雄市] $c 撰者 $d [民99]2010
215 0 $a 64面 $c 圖，表 $d 30公分
314 $a 指導教授：陳俞成博士
320 $a 參考書目：面
328 $a 碩士論文--國立高雄大學統計學研究所
330 $a 偏最小平方法(PLS)為多元線性迴歸、主成分分析與典型相關分析的綜合體，它可以處理小樣本以及存在共線性的資料。本論文主要比較近年來PLS2 演算法則之不同點。首先，我們介紹在本文中即將討論的幾種PLS2 與SIMPLS 的演算過程，並使用兩筆資料作實例演算。從實例演算的結果中可知，權重向量單位化與否對迴歸係數的估計會造成結果不同。我們修訂演算法的某些步驟，使它們合乎理論上的推導，最後，修訂之後的PLS 演算法再次使用同樣兩筆資料來進行比較，並以累積解釋變異來選取潛在變數之個數，以迴歸係數透過潛在變數解釋獨立變數與因變數的關係。 Partial least squares(PLS) is the complex of the multiple linear regression, principal component analysis and canonical correlation analysis. It can deal with small sample and the collinearity data. This paper is mainly the comparison of the PLS2 algorithm. First, we introduce several PLS2 process rules and SIMPLS discussed in this article, and use the two data for instance. By the examples, whether making a unit of weight vector or not will cause different to results of regression coefficient of different algorithms. We revised the algorithms to make them to be consistent with the theoretical derivation. After the revision of the PLS algorithm is used again to carry out the same two data sets in comparison. The cumulative explained variance is to select the number of factors, and the regression coefficients are to explain the relationship between independent variables and dependent variables through latent variables.
510 1 $a A Comparison of PLS2 Algorithms
610 0 $a 偏最小平方法 $a 小樣本資料 $a 共線性資料 $a PLS2 $a SIMPLS $a 迴歸係數 $a 累積解釋變異 $a 潛在變數
610 1 $a partial least squares $a small sample $a collinearity data $a PLS2 $a SIMPLS $a regression coefficient $a cumulative explained variance $a latent variables
681 $a 008M/0019 $b 343201 4421 $v 2007年版
700 1 $a 韓 $b 佩君 $4 撰 $3 483086
712 0 2 $a 國立高雄大學 $b 統計學研究所 $3 166081
801 0 $a tw $b 國立高雄大學 $c 20101103 $g CCR
856 7 $z 電子資源 $2 http $u http://handle.ncl.edu.tw/11296/ndltd/76968173275864377896