Perbandingan Kinerja Algoritma C.45 dan AHP-TOPSIS Sebagai Pendukung Keputusan Proses Seleksi Penerima Beasiswa

Abstract

Considering the importance of the scholarship, it is a challenge for the head of higher education institutions to make an accurate decision on giving scholarships and managing data of the recipients. However, the implementation is still less efficient in terms of time and allow for errors due to how to select and transfer of data recipients still use manual way. Decision support system is expected to improve performance and simplify the process of selection of recipients. In this research used data Bidik Misi scholarship. The input of this system is the personal data of applicants, the data of parents and academic data. Sample registries from 2013 - 2014 is used in this research. The data of applicants in 2013 are used for training. Whereas the data of applicants in 2014 are used for testing or target data. Input data will be processed using data mining techniques, C4.5 algorithm and Multiple Attribute Decision Making (MCDM) AHP-TOPSIS algorithm. The output of this system is a recommendation of applicants who are accepted and not accepted in Bidik Misi Scholarship and analysis algorithms based on performance.The test results for Bidik Misi Scholarship show that C4.5 algorithm got 85% on precision, 85% on recall and 93% on accuracy, and AHP-TOPSIS got 69% on precision, 69% on recall, and 87% on accuracy. In this research, C4.5 algorithm shows a better performance than AHP-TOPSIS algorithm.