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

Table 3 Average values of the classification metrics computed on the test set for the models achieving the highest F1-score in each of the problems

Problem	Best model	BA	F1	Precision	Sensitivity	Specificity
SCD vs cognitive impairment	RF	0.75	0.85	0.97	0.76	0.73
SCD vs ADD	XGB	0.84	0.92	0.94	0.90	0.79
SCD vs MCI	GA-SVM	0.69	0.84	0.93	0.77	0.62
MCI vs ADD	XGB	0.67	0.63	0.56	0.72	0.63

Abbreviations: BA, balanced accuracy; F1, f1-score; RF, random forest; XGB, extreme gradient boosting; GA-SVM, genetic algorithm-support vector machine; SCD, subjective cognitive decline; MCI, mild cognitive impairment; ADD, Alzheimer’s disease dementia

ISSN: 1758-9193