Autoencoders are an unsupervised learning model that aim to learn distributed representations of data.

Typically an autoencoder is a neural network trained to predict its own input data. A large enough network will simply memorize the training set, but there are a few things that can be done to generate useful distributed representations of input data, including:

1. constraining the size of the model, forcing it to learn a lower-dimensional representation that can be used to re-construct the original higher-dimensional data points.
2. adding artificial noise to the initial data points, and training the autoencoder to predict the data points minus the artificial noise. See denoising autoencoder for more information.