PCA

SVD

$X_{k \times } = U S V^T$

• $X = [x_1, x_2, …, x_m] \in R^{N \times m}$
  
• U is a m x m matrix, where $U^TU = I$
• V is a n x n matrix, where $V^TV = I$
• S is a m x n matrix, where.

Example

SVD fit 3D point

PCA