Fig. 2From: Compensation of feature selection biases accompanied with improved predictive performance for binary classification by using a novel ensemble feature selection approachPerformance of logistic regression models. On the y-axis the sensitivity and on the x-axis the specificity is shown. Three ROC curves are shown per dataset: of all features (solid), the AUC-FS selected (dashed) and the EFS selected (twodashed) features. The dotted line marks the performance of random guessingBack to article page