JMingKuo / fashion-mnist

Little experiments on fashion-mnist dataset with tensorflow.

fashion-mnist

Little experiments on fashion-mnist dataset with tensorflow. All results were produced without data preprocessing or augmentation and stopped after 100 epochs. The code of SELU activation function were borrowed from bioinf-jku's implementation.

The CNN architecture

C(3,32)-C(3,32)-P2-C(3,64)-C(3,64)-P2-FC64-FC64-S10

Results

Activation	Learning Rate	MNIST	Fashion MNIST
RELU	0.01	0.9874	0.8809
RELU	0.001	0.9388	0.7818
SELU	0.01	0.9871	0.8938
SELU	0.001	0.9490	0.8008

Curves

Fashion-MNIST (Loss)

Fashion-MNIST (Accuracy)

MNIST (Loss)

MNIST (Accuracy)