Table 2 Formulas for the performance metrics applied in this study.

From: Machine learning-based prediction of heating values in municipal solid waste

Index

Equation

Mean absolute error

\(MAE = \frac{{\mathop \sum \nolimits_{i = 1}^{N} \left| {act_{i} - pre_{i} } \right|}}{N}\quad \quad (2)\)

Mean square error

\(MSE = \frac{1}{N}\mathop \sum \limits_{i = 1}^{n} \left( {act_{i} - pre_{i} } \right)^{2} \quad \quad (3)\)

R-squared

\(R^{2} = \left( {\frac{{\mathop \sum \nolimits_{i = 1}^{n} \left( {act_{i} - \overline{act} } \right)\left( {pre_{i} - \overline{pre} } \right)}}{{\sqrt {\mathop \sum \nolimits_{i = 1}^{n} \left( {act_{i} - \overline{act} } \right)^{2} \mathop \sum \nolimits_{i = 1}^{n} \left( {pre_{i} - \overline{pre} } \right)^{{2{ }}} } }} } \right)^{2} \left( 4 \right)\)

Mean absolute percentage error

\(MAPE = \frac{1}{N}\mathop \sum \limits_{i = 1}^{n} \left| {\frac{{act_{i} - pre_{i} }}{{act_{i} }}} \right| \times 100 \quad \quad (5)\)

  1. where N indicates the number of observations; \(act_{i}\) and \(pre_{i}\) Illustrate the observed and predicted heating values, respectively; \(\overline{pre}\) Show the average of predicted heating values; \(\overline{act}\) Indicates the mean of the observed heating values.