Prediksi Nilai Mata Kuliah Mahasiswa Menggunakan Algoritma K-Apriori

Abstract

The main purpose of grade-course prediction is to help students to take elective courses corectly. The most students taking the course is based on the number of students taking the course. A set of student’s academic transcript can be analyzed for the patterns of association (association rule) between their subjects and grades. K-Apriori is one of data mining methods to find the patterns of their association rule to purpose for prediction of the other grade-course. The main step of this method is to cluster the data using K-Means and to get the pattern for the subjects and their grade using Apriori algorithm. How ever, there are missing values due to all offered courses are not taken by each students. Therefore, it is implemented preprocessing data using Wiener Transformation before it is applied to Apriori algorithm to find their pattern. The testing is based on student’s academic transcript using minimum support and confidence as 10% with lift ratio > 1. As a result, the generated rule from low and high GPA are achieved of error rate 8.75% and 8.5%. How ever, the generated rule from average GPA is achieved of error rate 11%