Prediksi Klasifikasi Royalti Batubara Menggunakan Algoritma Fuzzy K-Nearest Neighbor

Studi Kasus: CV.Fazar Utama

Abstract

As one of energy commodity, coal has been explored and exploited in order to fulfil the energy needs of world community. The calculation of coal royalties in Indonesia is based on the value of the quality of exploited or sold coal and the benchmark coal price determined by the government. CV Fazar Utama Kota Samarinda East Kalimantan Province classifies coal royalties based on calculations that have been made on coal sample inspection such as low, medium, and high. The Fuzzy K-Nearest Neighbor (FK-NN) method is one of the lazy learner algorithm which is used to determine the classification prediction. The K-Fold Cross Validation technique is used to obtain the optimal K value on the Fuzzy K-Nearest Neighbor (FK-NN) algorithm in order to get more accurate prediction. In this research, the first step is to find the optimal K value by experimenting 1-Fold Cross Validation, 3-Fold Cross Validation and 5-Fold Cross Validation using the 80:20 proportion of training and testing data. The next step is find the percentage accuracy of coal Royalty classification prediction in CV Fazar Utama in 2017 using Fuzzy K-Nearest Neighbor (FK-NN) Algorithm with optimal K value. Based on this research, the optimal value of K used in FK-NN Algorithm is F3-NN obtained in 1-Fold Cross Validation experiment. Furthermore, the result shows that the percentage accuracy of coal Royalty classification prediction at CV Fazar Utama in 2017 using F3-NN with 1-Fold Cross Validation is 100%.