Implementation of Orange Data Mining to Predict Student Graduation on Time at Pringsewu Muhammadiyah University

Abstract

Thel prolcelss olf molnitolring and elvaluating thel graduatioln olf Muhammadiyah Pringselwu Univelrsity (UMPRI) studelnts relally nelelds tol bel dolnel belcausel thel studelnt graduatioln ratel is an ellelmelnt olf accrelditatioln asselssmelnt that is velry impolrtant folr elach Study Prolgram. Data Mining can bel useld tol classify studelnt graduatioln accuracy. This study aims tol apply thel olrangel data mining applicatioln using thel K-Nelarelst Nelighbolr (K-NN), Delcisioln Trelel and Naivel Bayels moldells and will theln elvaluatel thel accuracy olf elach olf thelsel moldells. This relselarch was colnducteld at Pringselwu Muhammadiyah Univelrsity in selvelral batchels, theln studelnt data will bel analyzeld using thel olrangel data mining applicatioln using thel K-NN, Delcisioln Trelel and Naivel Bayels moldells. Thel data telsting prolcelss appliels K-Folld Crolss Validatioln (K=5), whilel thel elvaluatioln moldell useld is thel Colnfusioln Matrix and ROlC. Thel relsults olf thel colmparisoln olf thel threlel moldells arel as folllolws, K-NN has an accuracy ratel olf 75.7%, Delcisioln Trelel has an accuracy ratel olf 78.1%, and Naivel Bayels has an accuracy ratel olf 77.8%. Thelrelfolrel, folr classifying thel graduatioln ratel olf Muhammadiyah Univelrsity studelnts, Pringselwu relcolmmelnds thel Delcisioln Trelel moldell belcausel it has a belttelr lelvell olf accuracy than K-NN and Naivel Bayels.