Skip to main content

Advertisement

Table 1 Training set sizes tested during the study for the different methods and datasets

From: Large-scale ligand-based predictive modelling using support vector machines

Linear SVM RBF SVM
Solubility logD Solubility logD
100 100 100 100
1000 1000 1000 1000
5000 5000 5000 5000
20,000 20,000 20,000 20,000
32,096 40,000 32,096 40,000
  80,000   80,000
  160,000   160,000
  320,000   
  1,188,343