Figure 5.

Receiver operating characteristic curve (ROC) for c1, c2, and sample entropy. The area under the ROC curve (AUC) is shown in the legend. As classifier, logistic regression is used. In this case, c₁, c₂, and sample entropy, each 1st-3rd principal component was used separately. In the case represented by “ALL, all 1st–3rd principal components component of c₁, c₂, sample entropy were used. We evaluated ROC in the case using all these values; the results show the enhancement of classification ability (AUC = 1.00).