Table 3.

Validation and Test results of the Model B1 and Model C on the classification task.

Model	Val/Test	Class	Recall	Precision	F1 score	Accuracy
B1	5-fold CV	Non-AD	0.662 ± 0.175	0.670 ± 0.101	0.652 ± 0.125	0.662 ± 0.109
		AD	0.666 ± 0.170	0.675 ± 0.126	0.659 ± 0.139
	Test set	Non-AD	0.8333	0.6452	0.7272	0.6875
		AD	0.5416	0.7647	0.6341
C	5-fold CV	Non-AD	0.778 ± 0.104	0.673 ± 0.092	0.715 ± 0.070	0.693 ± 0.082
		AD	0.615 ± 0.151	0.743 ± 0.097	0.659 ± 0.112
	Test set	Non-AD	0.8333	0.6896	0.7547	0.7292
		AD	0.6250	0.7894	0.6976

Bold values represent the validation and test accuracies of best performing model amongst the models under consideration in the respective table.