METODE REGRESI GULUD UNTUK MENGATASI MASALAH MULTIKOLINEARITAS PADA KASUS INDEKS KUALITAS LINGKUNGAN HIDUP DI INDONESIA TAHUN 2021

Abstract

The environmental quality index is an indicator of environmental quality in Indonesia. EQI shows an increasing trend from 2018 to 2022, although Indonesia's environmental quality is ranked 116 out of 180 countries in the world. Therefore, the purpose of this study is to describe the characteristics of the EQI and the factors that can affect EQI, estimate the parameters of the ridge regression model and test the significance of the ridge regression model parameters to overcome the multicollinearity problem in the EQI case. This study uses 4 independent variables, namely population density, traffic, waste, and sanitation in Indonesia in 2021 as secondary data. In this study, ridge regression analysis is used because there is an almost linear relationship between the factors that are thought to affect the EQI. Ridge Regression is a technique that imposes limits (penalties) on parameter estimates in the regression model so that a reduction in the estimated coefficient value can overcome the multicollinearity problem. The results show that the optimal lambda value is 4.737078 when using the 5-fold cross-validation method. Variables that partially affect EQI are population density, waste, and environmental sanitation. The independent variables in this study that can explain the variability of EQI in Indonesia in 2021 is 59.52%, while 40.48% is explained by other variables that are not included in the model. Indonesia's EQI score in 2021 is 71.45 with the predicate "Good".

Keywords:   Environmental Quality Index, Ridge Regression, Cross Validation.