SIMULASI PENANGANAN MASALAH MULTIKOLINEARITAS DENGAN MENGGUNAKAN METODE PRINCIPAL COMPONENT REGRESSION (PCR) DAN METODE REGRESI RIDGE

HANNIFA IRMAJIHAN NABILLA, NIM 17106010042 (2021) SIMULASI PENANGANAN MASALAH MULTIKOLINEARITAS DENGAN MENGGUNAKAN METODE PRINCIPAL COMPONENT REGRESSION (PCR) DAN METODE REGRESI RIDGE. Skripsi thesis, UIN SUNAN KALIJAGA YOGYAKARTA.

Preview	Text (SIMULASI PENANGANAN MASALAH MULTIKOLINEARITAS DENGAN MENGGUNAKAN METODE PRINCIPAL COMPONENT REGRESSION (PCR) DAN METODE REGRESI RIDGE) 17106010042_BAB-I_IV-atau-V_DAFTAR-PUSTAKA.pdf - Published Version Download (2MB) | Preview
	Text (SIMULASI PENANGANAN MASALAH MULTIKOLINEARITAS DENGAN MENGGUNAKAN METODE PRINCIPAL COMPONENT REGRESSION (PCR) DAN METODE REGRESI RIDGE) 17106010042_BAB-II_sampai_SEBELUM-BAB-TERAKHIR.pdf - Published Version Restricted to Registered users only Download (5MB)

Abstract

Regression analysis is an analysis used to determine the effect between the independent variable and the dependent variable. One of the conditions that must be met in the classical assumption is that there is no multicollinearity between variables. Multicollinearity is a condition in which two or more independent variables have a perfect or almost perfect (collinear) linear relationship. The impact of multicollinearity is that although it is a BLUE (Best Linear Unbiased Estimator) but the Ordinary Least Square (OLS) estimator has a large variance and covariance so that it makes an accurate estimator difficult, the confidence interval tends to be wider, the t ratio of one or more coefficients tends to not significant, even though the 2 R value is high, the t ratio is only slightly significant, and the OLS estimator and its standard error are sensitive to small changes in the data. To overcome the problem of multicollinearity in this study the author uses two methods, namely the Principal Component Regression (PCR) method and the Ridge Regression method. PCR is a method that combines Principal Component Analysis (PCA) and Regression Analysis. While the Ridge Regression method is the development of the Ordinary Least Square (OLS) which produces a biased estimator of the regression coefficients. This study aims to determine the best method between the PCR method and the Ridge Regression method with the Mean Squared Error (MSE) criterion which has a minimum value. The MSE value is obtained by squaring the difference between the forecast value and the actual value. The PCR method and the Ridge Regression method were applied to the simulation data generated by the author from the R software with 30 replications. The variable used was the dependent variable Y involving five independent variables X1, X2 , X3, X4 ,  5 and X , where the variable 1 X and 5 X contained elements of multicollinearity. After analyzing the data, the average MSE value from the PCR model was 2487,69, and the average MSE value from the Ridge Regression model was 1036,92. Thus, in this simulation the Ridge Regression method is more effectively used to overcome the problem of multicollinearity than the PCR method.

Share this knowledge with your friends :

Actions (login required)

View Item