Skip to main content

Table 2 Statistics of best predictive models for AID449762

From: Computational models for in-vitro anti-tubercular activity of molecules based on high-throughput chemical biology screening datasets

Classifier* TP rate FP rate TN rate FN rate Accuracy ROC area BCR #
CSC NB 47.30 19.50 80.50 52.70 80.28% 0.70 63.90
CSC RF 47.00 19.20 80.80 53.00 80.58% 0.712 63.90
CSC SMO 51.90 19.30 80.70 48.10 80.52% 0.748 66.30
Metacost J48 40.60 19.10 80.90 59.40 80.62% 0.61 60.75
  1. *CSC denotes CostSensitiveClassifier, # BCR denotes Balanced Classification Rate