DETERMINATION OF CALORIE VALUE OF COAL BY USING MACHINE LEARNING METHODS

In this study, the caloric value of coal was estimated using three different machine learning methods. For this, sulfur, volatility and ash values of 1665 coal samples were used as independent variables for the determination of calories. Artificial neural networks, support vector regression and multiple linear regression methods were used as estimation methods. At the end of the study, according to the R2 performance measurement, thc most successful estimation methods were ANN (0.891), SVR (0.887) and MLR (0.882), respectively. According to RMSE analyzes, SVR (0.017), MLR (0.028) and ANN (0.029) were most successful, respectively. According to the MAE value; ANN (1.3%), SVR (1.7%) and MLR (1.8%) were most successful, respectively. The study can be considered successful when compared with similar studies in the literature.