Robust features for the automatic identification of autism spectrum disorder in children

Table 3 Classifier performance for various feature sets

Classifier	Our features	mMSE	Ours + mMSE
SVM	0.69	0.69	0.67
Logistic regression	0.67	0.63	0.63
Naive Bayes	0.77	0.60	0.79

The best performance for each set is in bold. Numbers shown denote the accuracy of the classifier. mMSE, modified multiscale entropy; SVM, support vector machine.

ISSN: 1866-1955