KAJIAN PENGGUNAAN REGRESI KOMPONEN UTAMA

A&rsquo (yunin Sofro, )

Matematika dan Sains
Vol 17, No 1 (2010)

Regression model should be followed many assumes, one of them is collinearity. Variance Inflation factor (VIF) is used to indentify multicollinear case and for handling is needed special analysis, one of them is stepwise regression model and Principal component regression. In this paper shows that the value of R2 from stepwise regression is 91,63% and with principal component regression is 89.8%. So, stepwise regression model better than principal component regression model for handling multicollinear.2 from stepwise regression is 91,63% and with principal component regression is 89.8%. So, stepwise regression model better than principal component regression model for handling multicollinear.