Choosing the number of components

Low rank factorization represents a data matrix using fewer dimensions. We consider singular value decomposition, principal component analysis, reconstruction, and component choice.

Available Monday, November 30, 2026.