Klasifikasi Rumah Tangga Miskin Di Kecamatan Kaubun Tahun 2020 Dengan Menggunakan Metode Improved Chi-Square Automatic Interaction Detection

Abstract

Classification was grouping samples based on similarities and differences by using categorical variables. The classification method used in this study was the Improved Chi-square Automatic Interaction Detection (I-CHAID) which was an improvement from the CHAID method. This research aims to provide an overview of poor households and classify poor households, and to compare the accuracy of the classification results for each data proportion. The data used is household data in Kaubun District in 2020 with poor and non-poor status. The results of this study indicate that there were 10 households with poor status, and households were classified as poor if the frequency of eating is less than 3 times a day, and the best classification accuracy results use the proportion of training data of 60% and testing data of 40%.