<![CDATA[Multicollinearity is a problem that must be addressed when using regression. Multicollinearity often occurs in socioeconomic data, such as Per Capita Expenditure. Several relevant studies have shown that Least Absolute Shrinkage and Selection Operator (LASSO) regression is a good method for handling multicollinearity. Additionally, it produces the simple model. Meanwhile, the Least Angle Regression (LAR) algorithm works effectively in model optimization, especially when multicollinearity occurs in multiple variables. Therefore, this study aims to handle multicollinearity with LAR LASSO regression in the specific case of per capita expenditure data in Wonosobo with many variables experiencing multicollinearity. The result study is LAR LASSO regression successfully eliminated two of the four predictor variables that exhibited multicollinearity by reducing the regression coefficients on the two predictor variables to zero. The best regression model obtained produces two significant coefficients so that Per Capita Expenditure in Wonosobo was influenced by the Human Development Index and Average Years of Schooling.]]>
Copyrights © 2025
