Fig. 2

30 features obtained by LASSO and XGBoost algorithms. a AUC with different number of characteristics as revealed by the LASSO model. b Misclassification error for different number of features revealed by the LASSO model. In a and b, the grey lines represent the standard error and the vertical dotted lines represent optimal values by minimum criteria (left) and the largest value of lambda such that the error is within one standard error of the minimum (right). The upper abscissa is the number of non-zero coefficients in the model at this time and the lower abscissa is log Lambda, which is the tuning parameter used for tenfold cross-validation in the LASSO model. c The intersection of non-zero coefficients in a and b. 80 non-zero coefficients are obtained in the LASSO model. d The best model features were ranked based on the gain index in xgboost model. The xgboost model further simplified the 80 features from the LASSO model, and finally, 30 valid features were obtained. The gain index represents the fractional contribution of each feature to the model based on the total gain of this feature’s splits