Simulasi Rancangan Pemetaan Sekolah dengan Metode Algoritma Machine Learning Menggunakan Software RapidMiner

<p><strong>The growth in the number of educational institutions creates competition which encourages each institution to have a special strategy deal with it. The challenge that must also be faced world education is uneven quality various regions. One way face equal distribution school create programs suit needs school. In this study, solving problem mapping schools owned by XYZ foundation was carried out. results study obtained design indicator instrument assess Then simulation from unsupervised learning K-Means and K-Medoids Clustering methods, as well predicting patterns supervised Decision Tree C4.5 method. were selected for proposed Davies Bouldin index value 0.112. model cluster K-Medoids, namely Excellent School, has 44 schools; at Good there are 36 Improvement 20 schools. Meanwhile, prediction pattern rules dominant attributes (IND) 1, 2, 3, 4. Also, using 80:20 ratio decision tree show new testing data assumption 101st goes cluster_1 an accuracy rate 95%.</strong></p><p><strong><em>Keywords</em></strong> - D<em>ecision tree,</em><strong> </strong><em>K-means, K-medoids, S</em><em>chool instrument.</em></p>