posted on 2020-12-23, 21:44authored byMaciej BuchowskiMaciej Buchowski, J. Dustin Tracy, Thomas Donnelly, Evan C. Sommer, William J. Heerman, Shari L. Barkin
From the group of
selected participant recordings
(n=400), 200 were randomly allocated to the development (train) group using the random sampling function in R. The development
group was used to optimize DT algorithm parameters values (block length,
threshold, bedrest-start
trigger, and bedrest-end trigger).