Skip to main content
. 2018 Sep 28;13(9):e0204644. doi: 10.1371/journal.pone.0204644

Fig 2. The graph shows values of the accuracy (percentage of correctly classified compounds) of the four classifier models over the 10 folds cross-validation.

Fig 2

The sequence minimization optimization (SMO) and Random Forest (RF) classifier models showed greater predictive accuracy than the Naïve Bayesian (NB) and Voted Perceptron (VP) classifier models.