Klasifikasi Batubara Berdasarkan Jenis Kalori Dengan Menggunakan Algoritma Modified K-Nearest Neighbor

Imalita Agustin (Laboratorium Statistika Komputasi FMIPA Universitas Mulawarman)
Yuki Novia Nasution (Laboratorium Matematika Komputasi FMIPA Universitas Mulawarman)
Wasono Wasono (Laboratorium Matematika Komputasi FMIPA Universitas Mulawarman)

Eksponensial
Vol 10 No 1 (2019)

Coal is a sedimentary rock containing the main elements Carbon (C), Hydrogen (H), and Oxygen (O). Examination of coal samples in the laboratory according to company operational standard based on Air Dried Basis (ADB) are the amount of water, ash content, flying substance, solid carbon, sulfur, and Gross Calorific Value. At PT. Pancaran Surya Abadi Anggana Subdistrict Kutai Kartanegara, coal is classified based on its calorie type namely Antrachite, Bituminous, and Sub-Bituminous. In this research Modified K-Nearest Neighbor (MKNN) Algorithm is used to predict the classification. The k-Fold Cross Validation technique is used to obtain the optimal K value on MKNN Algorithm for accuracy. A measurement based on this research, the K-Optimal value used in MKNN Algorithm for coal classification in PT.Pancaran Surya Abadi is 3-NN. The value of K = 3 produces the prediction accuracy of Coal Classification based on the type of calories in PT.Pancaran Surya Abadi on 100% testing data