PERBANDINGAN METODE K-NEAREST NEIGHBOR DAN FUZZY C-MEANS DALAM MENENTUKAN PREDIKAT KELULUSAN MAHASISWA

Abstrak

The research was purposed to compare the method performance for knowing the accuracy of K-Nearest Neighbor and Fuzzy C-Means in determining of students graduation in system information study program STMIK Palangka Raya. Extreme programming was used in this research because it was devoted to development to software appropriately. Some steps which used in research were planning by preparing the data and other necessary needs, designing, encoding on Matlab software, and testing. Based on the calculation of accuracy using MAPE for value the result of testing percentage in comparing the method obtained that Fuzzy C-Means method could predict by increasing the accuracy value along with increasing the training data,while the K-Nearest Neighbor method got the decreasing of accuracy value although the training data was in creased. But, the K-Nearest Neighbor method was still better on accuracy which the accuracy was above 90% and Fuzzy C-Means method had the accuracy below 90% in this research.