CD Skripsi
Perbandingan Metode Regresi Logistik Biner Dan Algoritma C5.0 Untuk Mengklasifikasi Wilayah Desa Perdesaan Dan Desa Perkotaan Di Kota Sungai Penuh
The classification of rural and urban areas has become an important aspect in development planning and policy-making. The aim of this study is to identify the best classification method and determine which variables significantly influence the classification of rural and urban villages in Sungai Penuh City. Binary logistic regression is used to identify the factors affecting area classification, while the C5.0 algorithm is applied to build a decision tree-based classification model. The results of the binary logistic regression analysis show that the variables significantly influencing area classification are population density and the availability of kindergartens (TK). The resulting model has an accuracy of 61.9% and an F1-score of 63.6%, indicating a level of classification accuracy that still has room for improvement. Meanwhile, the classification results using the C5.0 algorithm indicate that the variables that influence the differentiation between rural and urban villages are households with agricultural businesses (RTUP), population density, and the availability of kindergartens (TK). The model produced using the C5.0 algorithm has a higher accuracy of 76.2%, with an F1-score of 80%, indicating that the C5.0 method is more effective than binary logistic regression in classifying rural and urban village areas in Sungai Penuh City.
Keywords: Classification, rural and urban villages, Sungai Penuh City, binary logistic regression, C5.0 Algorithm.
Tidak tersedia versi lain