This study aims to identify and analyze the variables that affect the open unemployment rate in Banten Province, Indonesia. The analyzed variables include population density, average years of schooling, labor force participation rate, minimum wage, Provincial GRDP, total labor force, and the number of poor people. The method used in this study is multiple linear regression analysis with secondary data from the Central Bureau of Statistics (BPS) for the period 2017–2022. The analysis revealed multicollinearity in the average years of schooling variable, with a Variance Inflation Factor (VIF) >10. To address this issue, Lasso regression and Elastic Net regression were applied. The results of this study show that Lasso regression produces a model with a Mean Squared Error (MSE) of 1.3234857, while Elastic Net regression yields a model with a lower MSE of 0.180683, indicating better predictive performance. The best model for predicting the open unemployment rate in Banten Province is the Elastic Net regression. The variables that significantly affect the open unemployment rate are average years of schooling, labor force participation rate, minimum wage, Provincial GRDP, total labor force, and the number of poor people. The conclusion of this study is that Elastic Net regression is more effective in predicting the open unemployment rate than other methods. The implication of these findings is that the generated model can serve as a basis for formulating more effective labor policies to reduce the unemployment rate in Banten Province.