Using bundle embeddings to predict daily cortisol levels in human subjects

Table 6 Bootstrap distribution of coefficient 1 for the optimal models

	Full embedding		Night bundle		Day bundle		TOD corrected
ID	val	rmsd	val	rmsd	val	rmsd	val	rmsd
1	−1.06	0.10	−0.84	0.14	− 2.20	0.32	−1.18	0.23
2	−1.02	0.22	−0.82	0.16	−2.05	0.37	−1.38	0.34
3	−1.11	0.12	−0.94	0.05	−1.06	0.41	−1.28	0.23
4	−1.14	0.13	−0.69	0.10	−2.31	0.39	−1.28	0.25
5	−1.18	0.17	−1.03	0.05	−0.41	0.81	−1.13	0.17
6	−1.16	0.10	−0.90	0.15	−1.94	0.50	−1.32	0.28
7	−1.95	0.20	−1.03	0.14	−2.74	0.19	−1.74	0.30
8	−1.05	0.07	−0.84	0.09	−2.29	0.25	−1.33	0.18
9	−1.12	0.10	−1.29	0.31	−2.37	0.29	−1.54	0.31
10	−1.08	0.06	−0.94	0.10	−1.54	0.59	−1.60	0.39

Note: ID = subject ID, val = average value of coefficient 1, rmsd = root mean square deviation of the coefficient values relative to the average of the corresponding target

ISSN: 1471-2288