PENERAPAN MACHINE LEARNING REGRESI LOGISTIK UNTUK MEMPREDIKSI FAKTOR YANG MEMPENGARUHI INDEKS PRESTASI MAHASISWA

Abstract

This article aims to look at the factors that affect student Grade Point, which can come from internal and external sources. The method used for prediction of factors affecting student achievement index using logistic regression is CRISP-DM (Cross Industry Standard Process for Data Mining). The sample used was 96 Pendidikan Mathematics Students at UIN Sheikh Ali Hasan Ahmad Addary Padangsidimpuan through a questionnaire distributed using G-Form. The results found are the variables of working while studying, the number of hours of study, the cost of living per month and the relationship with friends have a significant effect on the students’ grade point average. The greatest influence on student GPA is the variable of study hours, followed by the variable of monthly living expenses, next is working while studying and last is the relationship with peers. The variables of working while studying, the number of hours of study, the cost of living per month and relationships with peers contributed an influence of 32.5%, the rest by other factors besides the factors in this article which were found to have no significant effect. The model built is good enough with an accuracy performance level of 78.9%.