Anormaly Detection With Deep Learning Encoder Decoder(https://arxiv.org/abs/1607.00148?context=cs)
This image is quote from page.2
you make directory like this
Enc-Dec-AD
- .gitignore
- EncDecAD.py
- README.md
- data
- anormaly_data
- train_and_test
- train.npy
- test.npy
- model
if you want to change the train source path and test source path,
you can change it when you make instance of EncDecAD class.
model = EncDecAD("train_source", "test_source")
- calculation of Gaussian params μ and Σ