06.09.2021 Views

Linear Algebra, Theory And Applications, 2012a

Linear Algebra, Theory And Applications, 2012a

Linear Algebra, Theory And Applications, 2012a

SHOW MORE
SHOW LESS

Create successful ePaper yourself

Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.

328 SELF ADJOINT OPERATORS<br />

where σ is of the form<br />

⎛<br />

⎞<br />

σ 1 0<br />

⎜<br />

σ = ⎝<br />

. ..<br />

⎟<br />

⎠<br />

0 σ k<br />

for the σ i the singular values of A, arranged in order of decreasing size.<br />

Proof: By the above lemma and Theorem 13.3.3 there exists an orthonormal basis,<br />

{v i } n i=1 such that A∗ Av i = σ 2 i v i where σ 2 i > 0fori =1, ··· ,k,(σ i > 0) , and equals zero if<br />

i>k.Thus for i>k,Av i = 0 because<br />

For i =1, ··· ,k, define u i ∈ F m by<br />

Thus Av i = σ i u i . Now<br />

(Av i ,Av i )=(A ∗ Av i , v i )=(0, v i )=0.<br />

(u i , u j ) = ( σ −1<br />

i<br />

= ( σ −1<br />

i<br />

u i ≡ σ −1<br />

i Av i .<br />

Av i ,σ −1<br />

j<br />

) (<br />

Av j = σ<br />

−1<br />

i v i ,σ −1<br />

j A ∗ )<br />

Av j<br />

v i ,σ −1<br />

j σ 2 ) σ j<br />

jv j = (v i , v j )=δ ij .<br />

σ i<br />

Thus {u i } k i=1 is an orthonormal set of vectors in Fm . Also,<br />

AA ∗ u i = AA ∗ σ −1<br />

i<br />

Av i = σ −1<br />

i AA ∗ Av i = σ −1<br />

i Aσ 2 i v i = σ 2 i u i .<br />

Now extend {u i } k i=1 to an orthonormal basis for all of Fm , {u i } m i=1<br />

and let<br />

U ≡ ( )<br />

u 1 ··· u m<br />

while<br />

V ≡ ( )<br />

v 1 ··· v n .<br />

Thus U is the matrix which has the u i as columns and V is defined as the matrix which has<br />

the v i as columns. Then<br />

⎛<br />

u ∗ ⎞<br />

1<br />

.<br />

U ∗ AV =<br />

u ∗ k<br />

A ( )<br />

v 1 ··· v n<br />

⎜ ⎟<br />

⎝<br />

. ⎠<br />

u ∗ m<br />

⎛<br />

=<br />

⎜<br />

⎝<br />

u ∗ 1<br />

.<br />

.<br />

u ∗ k<br />

.<br />

.<br />

u ∗ m<br />

⎞<br />

( σ1 u 1 ··· σ k u k 0 ··· 0 ) (<br />

σ 0<br />

=<br />

⎟<br />

0 0<br />

⎠<br />

where σ is given in the statement of the theorem. <br />

The singular value decomposition has as an immediate corollary the following interesting<br />

result.<br />

)

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!