Knowledge Discovery Data Akademik Untuk Prediksi Pengunduran Diri Calon Mahasiswa

Admission of new students to an educational institution such as STMIK STIKOM Bali was an activity which is routinely implemented every new academic year. The registration of new student candidates was always increasing from year to year, but not all prospective students continued registration step of a number of prospective students who had passed. It would be too late to take action if a new student enrolled very little. By not knowing the number of registration students, institution cannot measure the time and the number of new admissions target which had been achieved. In this case the use of data mining techniques was expected to provide knowledge or information that was previously hidden in the data warehouse, thus becoming valuable information for the organization or institution. In this study, the classification model and the frequent pattern are made to identify the data pattern and its appearance for the "advanced" or "backward registration" status class. Some task mining was used to predict the prospective student was by classification techniques and techniques Frequent Pattern which extracted model and describe important data classes. The algorithm used is Decision Tree. The software which was used for implementation is WEKA.   Index Term : Data Mining, Classification, Decision Tree, Frequent Pattern