PEMODELAN INDEKS KEBAHAGIAAN NEGARA DENGAN METODE MULTIVARIATE ADAPTIVE REGRESSION SPLINE

Abstract

The level of happiness in a region can be reflected by its happiness index, influenced by several factors such as Gross Domestic Product (GDP) per capita, social relationships, healthy life expectancy, generosity, freedom, and perception of corruption. The aim of this research is to analyze the factors influencing the happiness index of a country in the year 2020. The appropriate method used for this case is the nonparametric regression approach with Multivariate Adaptive Regression Splines (MARS). The study resulted in the best model with a combination of basis functions (BF) 12, 18, and 24, with input variables (MI) set at 1 and output variables (MO) at 3, yielding a Generalized Cross Validation (GCV) value of 0.367. The predictor variables tested have a significant simultaneous and partial impact on the country's happiness index, with social relationships and healthy life expectancy being notable contributors.