. 2020 Jul 3;10(7):427. doi: 10.3390/brainsci10070427

Table 5.

The training accuracy of ten optimizers on a proposed patch-wise model architecture of CNN.

Epoch →	50	100	150	200	250	300
Optimizers ↓	50	100	150	200	250	300
Adam	0.97	0.98	0.98	0.98	0.99	0.99
Adagrad	0.95	0.96	0.96	0.96	0.96	0.96
AdaDelta	0.95	0.96	0.96	0.96	0.96	0.96
SGD	0.95	0.967	0.968	0.97	0.97	0.97
NAG	0.94	0.94	0.94	0.94	0.95	0.95
Rmsprop	0.95	0.95	0.95	0.95	0.95	0.95
Momentum	0.96	0.96	0.97	0.97	0.974	0.97
Adamax	0.95	0.95	0.95	0.96	0.96	0.96
CLR	0.96	0.96	0.96	0.96	0.96	0.96
Nadam	0.96	0.96	0.96	0.96	0.97	0.97