ANALISIS REGRESI KOMPONEN UTAMA UNTUK MENGATASI MASALAH MULTIKOLINIERITAS DALAM ANALISIS REGRESI LINIER BERGANDA

Marcus, Gresyea L.
Wattimanela, Henry J.
Lesnussa, Yopi A.

BAREKENG: Jurnal Ilmu Matematika dan Terapan
Vol 6 No 1 (2012): BAREKENG : Jurnal Ilmu Matematika dan Terapan

The climate in Ambon, are influenced by sea climate and season climate, cause of this island arrounded by sea, it is make very high rainfall intensity. A very high collinearity between independent variables, make the estimate can not rely be ordinary least square method so it market with not real regretion coefficient and the collinearity. Collinearity can be detected by linier correlation coefficient between independent variables and also with VIF way. Regretion principal component analysis is used to remove collinearity and all of independent variable into model, this analysis is regretion analysis technique wher eare combinated with principal component analysis technique. The object of this analysis is to simplify the variable by overcast it dimension, we can do it removes the correlation between coefficient by transformation. Regresion can help to solve this case rainfall in Ambon on 2010. So the colinearity to independent variables can be overcome and then we can get the best regretion rutes.