Estimasi Parameter Model Geographically Weighted Ridge Regression pada Indikator Pengukuran Penanganan Stunting di Indonesia

Anggun Yuliarum Qur’ani
Made Ayu Dwi Octavanny
Ratna Sari Widiastuti

OKTAL : Jurnal Ilmu Komputer dan Sains
Vol 2 No 08 (2023): OKTAL : Jurnal Ilmu Komputer Dan Sains

Regression that is implemented on cross-sectional data and weighted by geographic space coordinates is called Geographically Weighted Regression (GWR). When local multicollinearity problem is detected in GWR model, Geographically Weighted Ridge Regression (GWRR) can include this problem. One of the government's main agendas is to accelerate the reduction of stunting in children under five.. The prevalence of stunting among children under five shows a decrease between 2020 and 2021. GWRR performs very well in handling local multicollinearity problems by looking at the almost perfect coefficient of determination ( ) of 99.99815%. From all provinces in Indonesia, the biggest factors that influence IKPS are KPS/KKS or Food Aid Recipients, Education Dimension, and Immunization.