Predicting cognitive scores from wearable-based digital physiological features using machine learning: data from a clinical trial in mild cognitive impairment

Table 6 Average models’ performance (Pearson r) in predicting intra-individual changes in NTB scores across varying feature selection criteria

Cross-validation method	Cognitive measure	Machine Learning algorithms
Cross-validation method	Cognitive measure	Elastic Net	Random Forest	XGBoost
Subject-based	Global Cognition	− 0.02	0.06	0.09
	Executive Function	0.20	0.54	0.39
	Processing Speed	0.15	0.11	0.10
	Memory Immediate	0.26	0.16	0.17
	Memory Delayed	− 0.19	− 0.11	− 0.22
Interval-based	Global Cognition	0.16	0.17	0.19
	Executive Function	0.62	0.74	0.63
	Processing Speed	0.32	0.32	0.36
	Memory Immediate	0.10	0.13	0.01
	Memory Delayed	0.06	0.05	− 0.08

ISSN: 1741-7015