This preview has intentionally blurred sections. Sign up to view the full version.
View Full Document
Unformatted text preview: CS205 Review Session #7 Notes Useful Decompositions Consider an n × n matrix A that is symmetric positive semidefinite and has a nullspace of dimension p < n . We wish to factor A into M ˜ AM where the columns of the n × ( p n ) matrix M form an orthonormal basis for col( A ) = row( A ). It may help to first consider a concrete example: A = 2 1 0 1 2 = 1 0 0 1 0 0 2 1 1 2 1 0 0 0 1 0 = M ˜ AM T This particular choice for ˜ A and M is not necessarily unique. We can also write: A = 2 1 0 1 2 = 1 √ 2 1 √ 2 1 √ 2 1 √ 2 1 0 0 3 " 1 √ 2 1 √ 2 1 √ 2 1 √ 2 # One way to construct this factorization in general is to compute the spectral decomposi tion A = QΛQ T of A and trim of some of the unnecessary columns of Q . We can show, however, that no matter what method we use to determine A = M ˜ AM T , if the columns of M form an orthonormal basis for the column space of A , ˜ A must have several desirable properties. In particular:several desirable properties....
View
Full
Document
This note was uploaded on 01/29/2008 for the course CS 205A taught by Professor Fedkiw during the Fall '07 term at Stanford.
 Fall '07
 Fedkiw

Click to edit the document details