Table 1 The statistical parameters of model prediction performance

Classification prediction models

True positive (TP)

Real label = 1 and predicted label = 1

Number of correctly classified positive results

True negative (TN)

Real label = 0 and predicted label = 0

Number of correctly classified negative results

False positive (FP)

Real label = 0 and predicted label = 1

Number of misclassified positive results

False negative (FN)

Real label = 1 and predicted label = 0

Number of misclassified negative results

Accuracy (ACC)

ACC = (TP + TN)/(TP + TN + FP + FN)

Overall prediction accuracy

Sensitivity (SE)

SE = TP/(TP + FN)

Prediction accuracy of the positive set

Specificity (SP)

SP = TN/(TN + FP)

Prediction accuracy of the negative set

Precision

Precision = TP/(TP + FP)

Efficiency of positive results prediction

Recall

Recall = TP/(TP + FN)

Coverage of positive results prediction

Index F (F1)

F1 = 2Precison * Recall/(Precision + Recall)

Evaluation of the comprehensive performance of the models

Receiver operating characteristic (ROC) curve area under the Roc curve (AUC)

The probability that a randomly chosen positive example is ranked higher than a randomly chosen negative example

The performance of the classification model as its discrimination threshold is varied

Regression prediction model

Squared correlation coefficient (Q²/R²)

Q²/R² = \(1 - \frac{{\mathop \sum \nolimits_{i = 1}^{m} \left( {y_{i} - \widehat{{y_{i} }}} \right)^{2} }}{{\mathop \sum \nolimits_{i = 1}^{m} \left( {y_{i} - \overline{y}_{i} } \right)^{2} }}\)

Squared correlation coefficient

Mean absolute error of cross validation (MAE)

MAE = \(\frac{1}{m}\mathop \sum \limits_{i = 1}^{m} \left( {y_{i} - \widehat{{y_{i} }}} \right)^{2}\)

Mean absolute error of cross validation

Root mean squared error (RMSE)

RMSE = \(\surd \left( {{\text{MAE}}} \right)\)

Root mean squared error