Table 4.

Cohorts (number of MCI:HC), optimal feature sets and models of the experiments EXP1, EXP2, EXP3.

Cohort (MCI:HC)	Feature set	Model
EXP1: NLP features
10:7	nonstop, dvrst, HS, MDD	N.A
EXP2: Keystroke features
11:12	B and R indices	N.A
EXP3: Fused features
10:5	Nonstop, dvrst, MDD	Just NLP
	B and R indices	Just Keys
	Probabilities of Just NLP & B indices	A
	Probabilities of Just Keys & dvrst, nonstop, MDD	B
	Probabilities of Just NLP & probabilities of Just Keys	C

N.A., not applicable.