
Description Data matrices in machine learning are not square, so they require a step beyond eigenvalues: The Singular Value Decomposition (SVD) expresses every matrix by its singular values and vectors. Slides Used in this Video: Slides 25 through 30 Instructor: Gilbert Strang