Classification of underdeveloped regions in Maluku using binary MARS

Abstract

Based on Presidential Regulation Number 131 of 2015 on Determine of Underdeveloped Regions in 2015-2019, Maluku ranks 4th out of 23 provinces. Maluku has 11 regions, 8 of them are classified as underdeveloped regions. Classification of underdeveloped regions can be done using statistical analysis, namely the Binary Multivariate Adaptive Regression Spline (MARS). So, specific objectives to be achieved in this study are to determine the best Binary MARS model for classification and to calculate the accuracy of the Binary MARS model for the classification of underdeveloped regions in Maluku. After obtaining the classification results, we find out of GCV value for the MARS Binary model was 0.155 and the R2 value is 0.897. This model provided 100% accuracy in classifying the underdeveloped regions in Maluku.