AN EXPLANATION OF THE USE OF PRINCIPAL-COMPONENTS ANALYSIS TO DETECT AND CORRECT FOR MULTICOLLINEARITY

被引：75

作者：

LAFI, SQ ^{[1
]}

KANEENE, JB ^{[1
]}

机构：

[1] MICHIGAN STATE UNIV,CTR VET MED A109,CTR POPULAT MED,E LANSING,MI 48824

来源：

PREVENTIVE VETERINARY MEDICINE | 1992年 / 13卷 / 04期

关键词：

D O I：

10.1016/0167-5877(92)90041-D

中图分类号：

S85 [动物医学（兽医学）];

学科分类号：

0906 ;

摘要：

Multicollinearity can be a serious statistical problem in data analysis in which the contribution of each individual risk factor is being evaluated. Symptoms, effects and techniques that are useful in detecting the presence of multicollinearity in a data base are discussed. The mathematical basis of the principal-components analysis technique for detecting, quantifying, and adjusting the regression coefficients for the effects of multicollinearity in a data base was demonstrated.

引用

页码：261 / 275

页数：15

共 29 条

[1]

AFIFI AA, 1984, COMPUTER AIDED MULTI, P120

[2]

[Anonymous], 1977, REGRESSION ANAL EXAM

[3]

[Anonymous], 1984, ADV ECONOMETRIC METH

[4]

BELSLEY DA, 1980, REGRESSION DIAGNOSTI, P10

[5]

Bigras-Poulin M., 1985, THESIS U GUELPH GUEL

[6] THE USE OF REGRESSION FOR DETECTING COMPETITION WITH MULTICOLLINEAR DATA [J].

CARNES, BA ;

SLADE, NA .

ECOLOGY, 1988, 69 (04) :1266-1274

[7]

Conniffe D., 1973, Statistician, V22, P181, DOI 10.2307/2986767

[8]

DILLON WR, 1984, MULTIVARIATE ANAL ME, P15

[9] MSE EVALUATION OF RIDGE ESTIMATORS BASED ON STOCHASTIC PRIOR INFORMATION [J].

FOMBY, TB .

COMMUNICATIONS IN STATISTICS PART A-THEORY AND METHODS, 1977, 6 (13) :1245-1258

[10]

FOMBY TB, 1978, COMMUN STAT A, V8, P477

← 1 2 3 →