A.I. Wiki

Interested in AI? Get tips & stories in your inbox:

Denoising Autoencoders

An autoencoder is a neural network used for dimensionality reduction; that is, for feature selection and extraction. Autoencoders with more hidden layers than inputs run the risk of learning the identity function – where the output simply equals the input – thereby becoming useless.

Denoising autoencoders are an extension of the basic autoencoder, and represent a stochastic version of it. Denoising autoencoders attempt to address identity-function risk by randomly corrupting input (i.e. introducing noise) that the autoencoder must then reconstruct, or denoise.

Parameters and Corruption level

The amount of noise to apply to the input takes the form of a percentage. Typically, 30 percent, or 0.3, is fine, but if you have very little data, you may want to consider adding more.

Input/initiating a denoising autoencoder

Setting up a single-thread denoising autoencoder is easy.

To create the machine, you simply instantiate an AutoEncoder and set the corruptionLevel, or noise, as you can see in the example below.

That’s how you set up a denoising autoencoder with one visible layer and one hidden layer using MNIST data. This net has a learning rate of 0.1, momentum of of 0.9, and utilizes reconstruction cross entropy as its loss function.

Next, we’ll show you a stacked denoising autoencoder, which is simply many denoising autoencoders strung together.

Interactive Demo

Learn to build AI applications using our interactive learning portal.