PENERAPAN ALGORITMA C4.5 UNTUK PREDIKSI TINGKAT KEPUASAN MAHASISWA TERHADAP SISTEM INFORMASI AKADEMIK SEKOLAH TINGGI KEGURUAN DAN ILMU PENDIDIKAN HAMZANWADI SELONG

M. Khairul Rizal, Suhartini ., Jagat Sugiantara

Abstract


Development of Information Technology (IT) has provided the means for universities to improve the quality of service to its academic community. IT supported information system that can provide added value to each college if it is designed to be an information system that is effective and efficient. Effective use of information systems and efficient indicates that the system can support the achievement of the vision and mission of the college [1].

Classification technique is one of data mining techniques including supervised learning. Supervised learning means the process of establishing a correspondence (function) using a training dataset, seen as a "past experience" from a model. The goal is to predict from a value (output) of a function to each new object (input) after completing the training process [2].

C4.5 is one method used to induce a decision tree that was discovered by J. Ross Quinlan. This algorithm is derived from the popular ID3 algorithm used in making the decision tree. C4.5 is suitable algorithms used to classify large amounts of data into a certain class of classes based on the pattern of the existing data [3].

From the results of tests performed, both confusion matrix and ROC Curve proven that the results of tests performed C4.5 algorithm has an accuracy value of 98.18%.

From the research that the role of the C4.5 algorithm is able to predict the level of student satisfaction of the Academic Information Systems College of Teacher Training and Education Hamzanwadi Selong.

Full Text:

PDF

Article Metrics

Abstract view : 102 times
PDF - 232 times

Refbacks

  • There are currently no refbacks.