Algortihm C4.5 in mapping the admission patterns of new Students in Engingeering Computer

Data Mining

Abstract

University of Harapan Medan is one of the private universities in North Sumatra which has computer-based study programs such as Informatics Engineering and Information Systems. Every year this college receives many registrations from students who have completed their education at the school stage. The large number of incoming student data makes it difficult for the admin to select new students who will register. In this study using the C4.5 algorithm data mining method to map the pattern of student admissions selection in the field of Engineering and computers. The attributes used are the average value of report cards (high, enough, low), basic academic ability tests (very high, high, medium, low, very low), basic computer knowledge tests (very high, high, enough, low, very low) and interviews (good, bad). Data mining is a mathematical calculation process that uses algorithms and requires large data. While the C4.5 algorithm is an algorithm that processes data by calculating entrophy and information gain, where after the calculation process is carried out, those who get the largest information gain value will become nodes and branches. This C4.5 algorithm will describe a decision tree that will form a pattern in student selection. The results of this study indicate that in mapping the selection pattern of interview attributes into level 1 nodes, the attributes of the basic computer knowledge test become the level 1 branch, the attributes of the basic academic ability test become the level 2 branch and the attribute average value of report cards becomes the level 3 branch.

GS Cited Analysis