svd-pca-gha

This repository explores the use of singular value decomposition (SVD), principal component analysis (PCA) and Generalised Hebbian Learning (GHA), also known as Sangers rule.

Useful links / readings:

• Optimal Unsupervised Learning in a Single-Layer Linear Feedforward Neural Network, by TERENCE D. SANGER
• https://cs231n.github.io/neural-networks-2/ - SVD dataanalysis on CIFAR tutorial.