Originally developed for detecting enemy airplanes and warships during the World War II, the receiver operating characteristic (ROC) has been widely used in the biomedical field since the 1970s in, for example, patient risk group classification, outcome prediction and disease diagnosis. Today, it has become the gold standard for evaluating/comparing the performance of a classifier(s).A ROC curve is a two-dimensional plot that illustrates how well a classifier system works as the discrimination cut-off value is changed over the range of the predictor variable. The x axis or independent variable is the false positive rate for the predictive test. The y axis or dependent variable is the true positive rate for the predictive test. Each point in ROC space is a true positive/false positive data pair for a discrimination cut-off value of the predictive test. If the probability distributions for the true positive and false positive are both known, a ROC curve can be plotted from the cumulative distribution function. In most real applications, a data sample will yield a single point in the ROC space for each choice of discrimination cut-off. A perfect result would be the point (0, 1) indicating 0% false positives and 100% true positives. The generation of the true positive and false positive rates requires that we have a gold standard method for identifying true positive and true negative cases. To better understand a ROC curve, we will need to review the contingency table or confusion matrix.

No abstract

No abstract

No abstract