A Machine-Learned Predictor of Colonic Polyps Based on Urinary Metabolomics
Figure 1
Performance of linear SVM classifier using all metabolites and clinical features. ROC curve and convex hull showing tradeoff between true positive rate (sensitivity) and false positive rate (1-specificity) for performance: (a) on the training data (resubstitution error) and (b) the testing (evaluation) data, both during cross-validation. The overall performance of the 3 tested fecal tests is also shown.
