Skip to main content
. Author manuscript; available in PMC: 2021 Mar 12.
Published in final edited form as: KDD. 2017 Aug;2017:787–795. doi: 10.1145/3097983.3098126

Table 3:

Scalability result in per epoch training time in second (the number of epochs needed). SDP stands for Sequential Diagnoses Prediction

Model SDP
(Sutter data)
SDP
(MIMIC-III)
HF prediction
(Sutter HF cohort)

GRAM 525s (39 epochs) 2s (11 epochs) 12s (7 epochs)
RNN 352s (24 epochs) 1s (6 epochs) 8s (5 epochs)