1Cademy - Singular value decomposition (SVD)

Learn Before

Methods of dimensionality reduction
Linear Algebra

Concept

Singular value decomposition (SVD)

Singular value decomposition (SVD) is an unsupervised dimensionality reduction technique used for feature extraction. The singular value decomposition of an m x n matrix M is a factorization of the form U $\sum$ M*. U is an m x m unitary matrix. $\sum$ is a m x n diagonal matrix. V is a v x v unitary matrix. Image link: www.wikiwand.com/en/Singular_value_decomposition