$\sigma$ singular value of $A$ implies $\sigma$ singular value of $A^T$

181 Views Asked by At

Let $\sigma>0$ be a singular value of $A$, then it is also a singular value of $A^T$. Prove or disprove.

I think this should be true because if we consider the singular value decomposition of $A$, we will have:

$$A=U\Sigma V^T,$$

where $\Sigma$ contains $\sigma$.

Thus $$A^T=V\Sigma^TU^T.$$

Now let $U':=V$, $\Sigma':=\Sigma^T$, $V' :=U$, then

$$A^T=U' \Sigma' V'^T$$

is a singular value decomposition of $A^T$, and $\Sigma'$ contains $\sigma$. Moreover, if $\sigma=0$ then if $A$ is not a square matrix, then $\sigma$ is either a singular value of $A$ or $A^T$, but not both.

Please let me know if this is correct, or if there's a fallacy in my proof.

1

There are 1 best solutions below

2
On BEST ANSWER

Assuming that you are working with real matrices. I agree up to the point where you address the case $\sigma >0$. (which is actually what the question is asking).

Consider an example where $\sigma=0$, Let $A=(0,0)$.

$0$ is a singular value for both $A$ and $A^T$.

Your initial argument should work regardless of whether $\sigma =0$ or $\sigma >0$.