Unveiling the sound of the cognitive status: Machine Learning-based speech analysis in the Alzheimer’s disease spectrum

Table 9 Average values of the classification metrics computed on the test set for the binary classification problems

Problem	Model	BA	F1	Sen	Spe	Pre
SCD vs altered	RF	0.749	0.854	0.764	0.733	0.967
	XGB	0.739	0.884	0.819	0.659	0.960
	GA-SVM	0.725	0.875	0.805	0.644	0.958
	GA-KNN	0.680	0.742	0.605	0.756	0.961
SCD vs ADD	RF	0.820	0.916	0.900	0.741	0.933
	XGB	0.842	0.920	0.898	0.785	0.944
	GA-SVM	0.830	0.893	0.844	0.815	0.948
	GA-KNN	0.734	0.774	0.661	0.807	0.932
SCD vs MCI	RF	0.692	0.840	0.770	0.615	0.924
	XGB	0.680	0.807	0.715	0.644	0.925
	GA-SVM	0.693	0.841	0.772	0.615	0.925
	GA-KNN	0.628	0.656	0.508	0.748	0.925
MCI vs ADD	RF	0.668	0.614	0.669	0.668	0.569
	XGB	0.670	0.625	0.715	0.625	0.555
	GA-SVM	0.635	0.595	0.704	0.567	0.516
	GA-KNN	0.564	0.532	0.654	0.475	0.449

Abbreviations: BA, balanced accuracy; F1, f1-score; RF, random forest; XGB, extreme gradient boosting; GA-SVM, genetic algorithm-support vector machine; GA-KNN, genetic algorithm-K-nearest neighbors

ISSN: 1758-9193