Memprediksi Daftar Ulang Mahasiswa Baru Menggunakan Algoritma Bayesian Classification Di Universitas XYZ

Abstract

This study analyzes the PMB data in 2016. The PMB process at XYZ University has several paths, namely Regular, Transfer, and Bidikmisi. When the PMB process difficulties are encountered one of them is the number of prospective students who do not register again is increasing. Based on the identification of problems and information from officers the most important thing is the number of prospective students who do not re-register is increasing every year, thus affecting the impact of the number of PMB receipts. This study resolves these problems using the Bayesian Classification Algorithm so that the opportunity to re-register students can be known earlier. Based on the research conducted, it was concluded that the TDU class was 55% greater than the 45% DU class. This means that prospective students who do not completely re-register all study programs are larger. It was explained that PBSJ study programs 100% no prospective students who re-register. The Economic Education study program still has the opportunity of 75% of prospective students who are DU. 100% Physics Education study program no prospective students register. The Civil Engineering study program has a 25% chance that prospective students will be DU. The Agribusiness Study Program has the opportunity of 75% of prospective students doing DU. The 90% Animal Husbandry Study Program has no chance for prospective students to register. The Psychology study program has the same opportunities as the Animal Husbandry study program which is 90%, there is no chance that prospective students register. For the Law study program, it is possible for 90% of prospective students to register.