PENERAPAN MODEL REGRESI SEMIPARAMETRIK DERET FOURIER UNTUK MENGIDENTIFIKASI FAKTOR PENENTU ANGKA HARAPAN HIDUP DALAM KONTEKS SDGS 3

Sifriyani Sifriyani (Program Studi Statistika, Jurusan Matematika, Fakultas Matematika dan Ilmu Pengetahuan Alam, Universitas Mulawarman)
Jesselin Paskalis Sitinjak (Program Studi Statistika, Jurusan Matematika, Fakultas Matematika dan Ilmu Pengetahuan Alam, Universitas Mulawarman)
Andrea Tri Rian Dani (Program Studi Statistika, Jurusan Matematika, Fakultas Matematika dan Ilmu Pengetahuan Alam, Universitas Mulawarman)

Jurnal Gaussian
Vol 14, No 2 (2025): Jurnal Gaussian

Semiparametric regression combines parametric and nonparametric regression, applied when some relationships between the response and predictor variables are known while others are unknown. This study employed linear regression for the parametric component and the Fourier series estimator for the nonparametric component. The data consisted of life expectancy in Indonesia and its influencing factors, relevant to Sustainable Development Goals (SDGs) 3. The relationship between life expectancy and some predictors (maternal mortality rate, poverty rate) was linear, whereas with others (open unemployment rate, average years of schooling, Gini ratio, stunting prevalence, exclusive breastfeeding) the pattern was unknown and tended to be periodic. This characteristic aligns with the strength of the Fourier series semiparametric regression, which combines linear modeling with Fourier series components for nonlinear periodic relationships. The objective was to determine the optimal number of oscillations using the Generalized Maximum Likelihood (GML) method, obtain the best model, and identify factors affecting life expectancy. Results showed the best model had three oscillations with a GML of , RMSEP of 0.99, and R² of 82.67%. Significant factors included poverty rate, open unemployment rate, average years of schooling, Gini ratio, stunting prevalence, and exclusive breastfeeding.