Autoencoders (Ordinary type)

the Algorithm returns a fully trained autoencoder based ELM, you can use it to train a deep network by changing the original feature representations,it code or decode any input simple depending on the training parameters (input and output weights ) .

please cite as :

B. Tarek, H. Mouss, O. Kadri, L. Saïdi, and M. Benbouzid, “Aircraft Engines Remaining Useful Life Prediction using an Improved Online Sequential Extreme Learning Machine,” Appl. Sci., 2020.

In this link an example of regenerating of an image from the encoded matrix using an autoencoder is illustrated:

https://www.youtube.com/watch?v=ZdyUnbbSdN8&feature=youtu.be