Examining the different factors affecting automobile ownership with regression models

The aim of the study is to estimate the ownership of automobiles per capita using variables such as population, Consumer Price Index (CPI), road length, GDP, and exchange rate of the dollar. For that purpose, multiple regression analysis is applied on the data set in this study. It was found that a multicollinearity problem has occurred in the analysis performed. Regression methods from alternative regression methods such as Ridge, Lasso, and Elastic-Net are used to solve this problem. It has been found that the coefficient of determination for Ridge, Lasso, and Elastic-Net regression methods are 0.953, 0.976, and 0.952, respectively; adjusted coefficients of determination are 0.932, 0.972, and 0.943, respectively. The mean squared errors for the same methods are found to be 6.044, 4.461, and 4.655, respectively. According to the results obtained, it is seen that the Lasso regression method is the most suitable method with the smallest mean squared error and the largest coefficient of determination from the alternative models.