A Deep Learning Model for Stroke Patients’ Motor Function Prediction

Table 7

Floating-point operations per second and training CPU time of training of all methods.

Method	Floating-point operations per second (millions per second)	Average minutes	Standard deviation
EEG-DenseNet	0.052	142	±8.9
DenseNet	2.479	470	±8.7
Xception	6.243	409	±10.8
ResNet	3.965	616	±12.9
VGG16	13.15	855	±14.6