https://autoencoder-anomaly-detection.herokuapp.com/
Autoencoders are deep neural networks that learn to predict compressed versions of their inputs. model.fit(x_train, x_train)
The number of neurons in the middle layers are usually set to be fewer than the number of input neurons, so that the network learns just the essential information in the inputs.
Application areas include Anomaly Detection, Image Denoising, etc.