Skip to main content
. Author manuscript; available in PMC: 2024 Mar 1.
Published in final edited form as: Pac Symp Biocomput. 2024;29:327–340.

Table 1.

Size of sets used for training and evaluation for each population. Note that a combination of white British as the source domain plus another population as the target domain is always used.

Population Training labeled Training unlabeled Val. + Test labeled Total
White British * 0 0 *
Nigeria 10 213 106 + 107 852
Sri Lanka 10 134 67+ 67 535
Hawaii 822 5,862 412 + 413 7,507
*

White British set size depends on class balancing, but in all cases is >40,000.