Skip to main content

Table 2 Performance comparison of LightGBM with other machine learning methods

From: PDRLGB: precise DNA-binding residue prediction using a light gradient boosting machine

Dataset Methods ACC SN SP ST PRE F1 MCC AUC
PDNA-62 SVM 0.817 0.745 0.829 0.787 0.433 0.547 0.468 0.873
  Adaboost 0.814 0.791 0.818 0.804 0.431 0.558 0.485 0.881
  RF 0.817 0.782 0.822 0.802 0.435 0.559 0.486 0.883
  LightGBM 0.815 0.863 0.806 0.835 0.438 0.581 0.523 0.912
PDNA-224 SVM 0.786 0.765 0.776 0.771 0.194 0.310 0.306 0.847
  Adaboost 0.773 0.761 0.774 0.768 0.192 0.307 0.320 0.851
  RF 0.814 0.750 0.819 0.784 0.226 0.347 0.351 0.864
  LightGBM 0.800 0.833 0.797 0.815 0.224 0.353 0.383 0.896