From: Evaluation of maxout activations in deep learning across several big data domains
Activation | Average accuracy (%) | Accuracy HSD | Average 100 batches time (s) | Average 100 batches training time |
---|---|---|---|---|
ReLU 2x | 90.41 | a | 15.57 | 594.15 |
Maxout 3-2 | 90.35 | a | 62.67 | 1485.35 |
ReLU | 90.26 | ab | 7.41 | 349.52 |
Maxout 3-1 | 90.19 | ab | 29.67 | 972.79 |
Maxout 2-1 | 89.97 | ab | 17.50 | 440.28 |
Maxout 6-1 | 89.89 | ab | 48.83 | 1866.42 |
LReLU | 89.71 | b | 13.86 | 754.60 |
Tanh | 87.57 | c | 7.45 | 242.19 |
SeLU | 83.81 | d | 12.20 | 229.62 |