skip to main content

### ESIMASI PARAMETER REGRESI RIDGE MENGGUNAKAN ITERASI HOERL, KENNARD, DAN BALDWIN (HKB) UNTUK PENANGANAN MULTIKOLINIERITAS

Citation Format:
Abstract

Regression analysis is statistical method used to analyze the dependence of respond variables to predictor variable. In multiple linear regression analysis, there are assumptions that must be met, they are normality, homoscedasticity, absence of multicollinearity, and absence of autocorrelation. One of assumption frequently found is multicollinearity. If multicollineraity is exist between predictor variables, then regression analysis with ordinary least square is no longer used. Ridge regression is regression method to handle multicollinearity. The ridge estimator involves adding biasing constant (k) to each diagonal element of  X’X. Biasing constant (k) is determined by Hoerl, Kennard, and Baldwin (HKB) iteration method. This regression can be applied to inflation rate in Indonesia data and the factors that influence, they are BI rate, money supply, and exchange rate of rupiah. Ridge regression analysis, the VIF (Variance Inflation Factor) values for each predictor variables BI rate, money supply, and exchange rate of rupiah are 1.6637, 3.2712, and 4.3309. Since
VIF values are not exceed to 10, then there is no multicollinearity in ridge regression model.

Keywords: Inflation,  Multikollinearity, Ridge Regression,  HKB Iteration, VIF

Fulltext View|Download
Keywords: Inflation, Multikollinearity, Ridge Regression, HKB Iteration, VIF

Article Metrics:

Article Info
Section: Articles
Recent articles

Last update:

No citation recorded.

Last update:

No citation recorded.