Hands-On Machine Learning on Google Cloud Platform

Autoencoder with Keras

As we said previously, an autoencoder is a neural network whose purpose is to code its input into small dimensions and the result obtained to be able to reconstruct the input itself. Autoencoders are made up of the union of the following two subnets: encoder and decoder. To these functions is added another; it's a loss function calculated as the distance between the amount of information loss between the compressed representation of the data and the decompressed representation. The encoder and the decoder will be differentiable with respect to the distance function, so the parameters of the encoding/decoding functions can be optimized to minimize the loss of reconstruction, using the gradient stochastic.