How would you apply ROC curves to a machine learning (ML) problem? a. We want to compare your ML diagnosis, ROC(3) with the performance by a radiologist, ROC(1), and a older system that is available at the moment at the clinic, ROC(2). The comparison is done using ROC curves. Explain the curves, and discuss the results. 1.0 0.90 0.75 .. ROC (D ROC (3) 0.5 ROc (2) 0.0 0.3 1,0 False Positive Fraction (FPF)