CSC2515:

Lecture 7 (post)
Independent Components Analysis,
and Autoencoders

Factor Analysis

A degeneracy in Factor Analysis

What structure does FA capture?

Using a non-Gaussian prior

The square, noise-free case

The ICA density model

The information maximization view of ICA

Overcomplete ICA

Self-supervised backpropagation

Self-supervised backprop and PCA

Self-supervised backprop in deep autoencoders

A Deep Autoencoder
(Ruslan Salakhutdinov)

A comparison of methods for compressing digit images to 30 real numbers.

Do the 30-D codes found by the deep autoencoder preserve the class structure of the data?

Slide 16