Second order approximation necessary zero gradient

Info iconThis preview shows page 1. Sign up to view the full content.

View Full Document Right Arrow Icon
This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: cient for min: Hessian positive definite. Sufficient for max: Hessian negative definite. × × × ∗ × × 11 ∗ × = singular values of the matrix . 12 Special Cases (full rank) Pseudoinverse = ∗ × × × Σ × × × × ∗ # , × × × Σ × = , ,…, ≥⋯≥ ≥ = ∗, =∗ Also called generalized inverse. Can be defined for rectangular and for singular matrices. Least Squares ∗ × # × # × # × ∗ ∗ × ∗ × # ∗ # × ∗ Overdetermined: no exact solution. Best solution in least-squares sense. Solution of le...
View Full Document

Ask a homework question - tutors are online