Classification of domestic and imported coal in Turkey by machine learning methods


Kayakuş M.

International Journal of Oil, Gas and Coal Technology, cilt.30, sa.3, ss.300-311, 2022 (SCI-Expanded) identifier

  • Yayın Türü: Makale / Tam Makale
  • Cilt numarası: 30 Sayı: 3
  • Basım Tarihi: 2022
  • Doi Numarası: 10.1504/ijogct.2022.123200
  • Dergi Adı: International Journal of Oil, Gas and Coal Technology
  • Derginin Tarandığı İndeksler: Science Citation Index Expanded (SCI-EXPANDED), Scopus, PASCAL, Compendex, INSPEC
  • Sayfa Sayıları: ss.300-311
  • Anahtar Kelimeler: classification, domestic coal, imported coal, machine learning, Turkey
  • Akdeniz Üniversitesi Adresli: Evet

Özet

Copyright © 2022 Inderscience Enterprises Ltd.Coal is an energy source with high economic value. Although Turkey has its own coal reserves, it remains inadequate due to the increasing energy demand. Hence, it imports a significant amount of coal from other countries. There are significant differences in the tonnage prices of imported and domestic coal in Turkey. Imported coal is approximately four times more expensive than domestic coal due to the quality. It is important to classify coals as imported and domestic due to their pecuniary value. In this study, considering the calorie, sulphur, volatility and ash values of 1,665 coal samples, coal was classified into two categories as imported and domestic. Support vector machines from machine learning methods, decision trees and Bayes theorem are used for classification processes. As a result of the classification made with decision trees, the accuracy is 99.7%; 99.099% with support vector classification and 96.997% according to Bayes theory.