Table 1.

The list of feature sets with the number of features and data samples used in the machine learning pipeline after handling missing values and the number of selected features during the cross-validation process.

Feature set	Number of features	Number of samples	Number of features selected during cross-validation process
			LASSOa		NRLRb
			In all folds	In at least one fold	In all folds	In at least one fold
Bluetooth	3201	115	203	1026	278	1864
Calls	605	108	30	134	34	142
Campus map	16,381	111	66	455	12	161
Location	10,237	106	345	784	14	124
Screen	15,446	113	96	467	8	52
Sleep	5889	107	87	534	23	266
Steps	3055	107	270	485	0	8
Average	7831	110	157	555	53	374

^aLASSO: least absolute shrinkage and selection operator.

^bNRLR: nested randomized logistic regression.