Sistem Pendukung Keputusan Seleksi Penerimaan Calon Siswa/i Baru Menggunakan Algoritma C4.5 (Studi Kasus: SDIT An-Najah Jatinom Klaten)

Jurnal Teknoin Volume 21, Nomor 1, Maret 2015 ISSN : 0853-8697, Hal. 7-22


Abstract :

Education is very important and one of the keys of the progress of a country. Basic education or elementary school is the first stage of a long journey for a student in taking his or her education. The more parents has awareness in the importance of an education so the more the parents will enroll their children to study in school. Although the number of schools in a certain place is already considered enough, admission of new students still become a unique phenomenon because number of prospective new students entering school is huge but the admission method still uses manual techniques. Whether using a test or previous level report book, the rank and decision making about the acceptance of the students is still done manually. It can also provide opportunities for certain parties to commit fraud in the admission selection from subjective point of view (outside of the fixed criteria) which makes incorrect calculation and grading so the results are not properly valid.

This research would apply the rules that is formed from the decision trees on admission of new students selection process at SDIT An-Najah Jatinom Klaten. A decision tree consists of a set of rules for dividing a heterogeneous population into smaller parts, and more homogeneous with noticing to the goal variables. C4.5 algorithm is one of the algorithms used to form a decision tree.

The purpose of this research is to apply the rules that are formed from the C4.5 decision tree algorithm in admission of new students selection process. The registration (log) is expected to be more structured through this system so operator can easily and quickly process the calculation and ranking the prospective new student. Service becomes more rapid, accurate, effective, and efficient in terms of time and place because the piled print files (takes place and time) can be concise (stored in the database).

The results shows that the system is able to provide recommendations for the admission process based on the criteria and history that have ever happened. Based on sample data of 100 data among several criterias (religion, age, the average time from the each student’ place to the school, parents income per month, number of siblings)is is found that religious criteria have the highest gain value in the calculation of the initial node, ie by 0.367877. Finnaly, this system has been running correctly and can help the school admission committee in the selection of new candidate students.

Keywords: Decision Support Systems, Algorithms C.45, Gain.

Full text : Download PDF  (191)

Back ...