Figure 1
From: Improving classification in protein structure databases using text mining
Test ROC curves of the text similarity algorithm in the 'borderline' cases dataset. ROC curves of the test set from the 'borderline' cases DC1.1993 dataset for the TEXT (green), SSAP (black) classifiers and the logistic regression model that includes SSAP and TEXT independent variables (blue). The reference set was textCATH. The inset shows the same curves for low error rates (FPR < 0.10).