Anggita Nariswari
Department of Mathematics, Universitas Airlangga, Surabaya, Indonesia

Published : 1 Documents Claim Missing Document
Claim Missing Document
Check
Articles

Found 1 Documents
Search

Determinants of the Village Development Index (IDM) on Madura Island: A Classical and Machine Learning Comparison Dita Amelia; Addina Nurkamila; Anggita Nariswari; Fadli Akbar Pambudi
The Journal of Indonesia Sustainable Development Planning Vol 7 No 2 (2026): August
Publisher : Pusbindiklatren Bappenas

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.46456/jisdep.v7i2.1036

Abstract

Development disparities on Madura Island are characterized by an uneven distribution of basic social infrastructure, reflected in persistently low human development indicators and high poverty rates relative to other regencies in East Java. This study aims to identify fiscal and social infrastructure determinants of IDM status and to compare the performance of ordinal logistic regression (OLR), K-Nearest Neighbors (K-NN), and Support Vector Machine (SVM) in classifying IDM status across 62 sub-districts on Madura Island using 2024 data, evaluated under Stratified 10-Fold Cross Validation. K-NN achieved the highest overall accuracy (71.00%), followed by OLR (67.62%) and SVM (50.00%); however, the macro-averaged F1-Score of 0.42 for both K-NN and SVM indicates limited performance on minority classes. OLR reveals that healthcare facilities and equitable school access are significantly associated with IDM status. The models are best positioned as early-screening tools to support more responsive village development policies, with findings relevant to SDGs 1, 8, and 16.