Fig. 4From: Comparative analysis of machine learning algorithms for multi-syndrome classification of neurodegenerative syndromesConfusion matrix for every classification model. The values are averaged over all folds and figured as a percentage value of the true label. The coloration depends on the number of predicted cases in percent (the scale goes from the highest value (dark blue) to the value 0 (white)). The confusion matrix shows row normalized percentages, which results in precision are shown for matching classes in column and rowBack to article page