hw4_2011_04_22_02_solutions

# hw4_2011_04_22_02_solutions - EE263 S Lall 2011.04.22.02...

This preview shows pages 1–2. Sign up to view the full content.

This preview has intentionally blurred sections. Sign up to view the full version.

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

Unformatted text preview: EE263 S. Lall 2011.04.22.02 Homework 4 Solutions Due Thursday 4/28. 1. Some basic properties of eigenvalues. 50045 Show that (a) the eigenvalues of A and A T are the same (b) A is invertible if and only if A does not have a zero eigenvalue (c) if the eigenvalues of A are λ 1 ,...,λ n and A is invertible, then the eigenvalues of A- 1 are 1 /λ 1 ,..., 1 /λ n , (d) the eigenvalues of A and T- 1 AT are the same. Hint: you’ll need to use the facts that det A = det( A T ), det( AB ) = det A det B , and, if A is invertible, det A- 1 = 1 / det A . Solution. (a) The eigenvalues of a matrix A are given by the roots of the polynomial det( sI − A ). From determinant properties we know that det( sI − A ) = det( sI − A ) T = det( sI − A T ). We conclude that the eigenvalues of A and A T are the same. (b) First we recall that A is invertible if and only if det( A ) negationslash = 0. But det( A ) negationslash = 0 ⇐⇒ det( − A ) negationslash = 0. i. If 0 is an eigenvalue of A , then det( sI − A ) = 0 when s = 0. It follows that det( − A ) = 0 and thus det( A ) = 0, and A is not invertible. From this fact we conclude that if A is invertible, then 0 is not an eigenvalue of A . ii. If A is not invertible, then det( A ) = det( − A ) = 0. This means that, for s = 0, det( sI − A ) = 0, and we conclude that in this case 0 must be an eigenvalue of A . From this fact it follows that if 0 is not an eigenvalue of A , then A is invertible. (c) From the results of the last item we see that 0 is not an eigenvalue of A . Now consider the eigenvalue/eigenvector pair ( λ i ,x i ) of A . This pair satisfies Ax i = λ i x i . Now, since A is invertible, λ i is invertible. Multiplying both sides by A- 1 and λ- 1 i we have λ- 1 i x i = A- 1 x i , and from this we conclude that the eigenvalues of the inverse are the inverse of the eigenvalues. (d) First we note that det( sI − A ) = det( I ( sI − A )) = det( T- 1 T ( sI − A )). Now, from determinant properties, we have det( T- 1 T ( sI − A )) = det( T- 1 ( sI − A ) T ). But this is also equal to det( sI − T- 1 AT ), and the conclusion is that the eigenvalues of A and T- 1 AT are the same. 2. Norm expressions for quadratic forms. 55020 Let f ( x ) = x T Ax (with A = A T ∈ R n × n ) be a quadratic form. (a) Show that f is positive semidefinite ( i.e. , A ≥ 0) if and only if it can be expressed as f ( x ) = bardbl Fx bardbl 2 for some matrix F ∈ R k × n . Explain how to find such an F (when A ≥ 0). What is the size of the smallest such F ( i.e. , how small can k be)? (b) Show that f can be expressed as a difference of squared norms, in the form f ( x ) = bardbl Fx bardbl 2 − bardbl Gx bardbl 2 , for some appropriate matrices F and G . How small can the sizes of F and G be?...
View Full Document

{[ snackBarMessage ]}

### Page1 / 8

hw4_2011_04_22_02_solutions - EE263 S Lall 2011.04.22.02...

This preview shows document pages 1 - 2. Sign up to view the full document.

View Full Document
Ask a homework question - tutors are online