International Journal of Academic and Applied Research (IJAAR)   Year: 2023 | Volume: 7 | Issue: 11 | Page No.: 32-35 Modeling the Open Unemployment Rate in Indonesia Using Multivariate Adaptive Regression Spline Download PDF Andika Priyatama, Ardi Kurniawan, Elly Ana, M. Fariz Fadillah M, Sediono Abstract: The Open Unemployment Rate is the percentage of the number of unemployed to the total labor force. The open unemployment rate is an indicator used to measure labor that is not absorbed by the labor market and illustrates the underutilization of labor supply. The unemployment rate for the Indonesian workforce aged 15-24 years in 2021 is in second place after Brunei Darussalam in Southeast Asia. The aim of this research focuses on describing and modeling the level of open unemployment rate in Indonesia, and interpreting the best model results obtained. The method used is a method with a nonparametric regression approach, namely the Multivariate Adaptive Regression Spline. The results showed that the best model obtained was in a combination of 14 basis functions, 2 maximum interactions, and 1 minimum observation between knots. From this model, the predictor variables that have the most influence on the response variable in order based on the level of variable importance are education level, population density, Foreign Direct Investment (FDI), percentage of poor population, high school APK, economic growth rate, and gini ratio. In addition, there is also an interaction between two variables, namely education level with population density and FDI with population density. The best MARS model for open unemployment rate data in Indonesia produces a Generalized Cross Validation value of 1.915, R2 of 0.740, and Mean Square Error of 0.979.