Figure 1.

Performance of RPISeq classifiers in predicting RPIs. Receiver operating characteristic (ROC) curves for RPI predictions, illustrating the trade-off between true positive rate and false positive rate for RPISeq-RF (random forest) and RPISeq-SVM (support vector machine) classifiers, using two datasets, RPI2241 and RPI369. The area under the curve (AUC) of each ROC is shown next to the curve. The AUC for a perfect classifier is 1, and for a random classifier = 0.5.

Muppirala et al. BMC Bioinformatics 2011 12:489   doi:10.1186/1471-2105-12-489
Download authors' original image