Skip to main content

Table 5 Classification performance of the random forest on domains consisting of four, five and six SSEs in ten-fold cross-validation.

From: Automatic structure classification of small proteins using random forest

Shared SCOP Level 4SSEs 5SSEs 6SSEs
  Accuracy = 98% Accuracy = 98% Accuracy = 97%
  Pre Rec MCC Pre Rec MCC Pre Rec MCC
Class 0.99 0.99 0.92 0.98 1.00 0.89 0.97 1.00 0.85
Fold 0.96 0.83 0.89 1.00 0.69 0.82 0.95 0.51 0.70
Super-family 0.88 0.69 0.78 0.98 0.65 0.79 0.95 0.57 0.74
Family 0.98 0.92 0.95 0.98 0.92 0.94 0.98 0.84 0.90
  1. Classification performance of the random forest on domains consisting of four, five and six SSEs in ten-fold cross-validation. Pre = Precision, Rec = Recall and MCC = Matthew's correlation coefficient.