Skip to main content
. Author manuscript; available in PMC: 2019 Jan 30.
Published in final edited form as: Proc Int Conf Mach Learn Appl. 2019 Jan 17;2018:194–201. doi: 10.1109/ICMLA.2018.00036

TABLE II:

Training dataset counts for QAMat training

dataset relevant irrelevant
train (2014–2015) 23,466 23,466
test (2016) 16,706 16,706
(a) datasets for testing on year 2016
dataset relevant irrelevant
train (2014–2016) 33,075 33,075
test (2017) 9,582 9,582
(b) datasets for testing on year 2017