Table 3 Performance of combining different classifiers in the first layer
From: 6mA-StackingCV: an improved stacking ensemble model for predicting DNA N6-methyladenine site
First- layer classifiers | ACC (95%CI) | AUC (95%CI) |
|---|---|---|
XGBoost, Gradient boosting, LightGBM, random forest, logistic regression, decision tree | 0.9484 (0.9440–0.9528) | 0.9588 (0.9551–0.9624) |
XGBoost, Gradient boosting, LightGBM, random forest, logistic regression | 0.9486 (0.9444–0.9527) | 0.9589 (0.9557–0.9622) |
XGBoost, Gradient boosting, LightGBM, random forest | 0.9487 (0.9446–0.9529) | 0.9605 (0.9573–0.9637) |
XGBoost, Gradient boosting, LightGBM | 0.9488 (0.9445–0.9531) | 0.9616 (0.9586–0.9646) |
XGBoost, Gradient boosting | 0.9469 (0.9430–0.9509) | 0.9582 (0.9545–0.9620) |
XGBoost, LightGBM | 0. 9477 (0.9438–0.9517) | 0.9605 (0.9583–0.9627) |
Gradient boosting, LightGBM | 0.9449 (0.9409–0.9490) | 0.9597 (0.9572–0.9621) |