1
$\begingroup$

How can I prove that $\Vert AB \Vert_\infty\le\Vert A \Vert_\infty.\Vert B \Vert_\infty$ ?

What I have already done:

$\max_{1\le i \le n}(\sum_{j=1}^n|\sum_{k = 1}^n A_{ik}.B_{kj}|)$

$\le \max_{1\le i \le n}(\sum_{j=1}^n\sum_{k = 1}^n|A_{ik}.B_{kj}|)$

$=\max_{1\le i \le n}(\sum_{j=1}^n\sum_{k = 1}^n|A_{ik}|.|B_{kj}|)$

$=\max_{1\le i \le n}(\sum_{k=1}^n(|A_{ik}|.\sum_{j=1}^n|B_{kj}|))$

$\endgroup$
  • 1
    $\begingroup$ Have you tried to use the definition of the induced matrix norm? $\endgroup$ – EuklidAlexandria Mar 19 at 23:03
  • 1
    $\begingroup$ Please show the steps that you have tried first. $\endgroup$ – rash Mar 19 at 23:19
  • $\begingroup$ @rash That is what I have done until now. I think that I am almost getting there but I am stuck. $\endgroup$ – Rebeca Silva Mar 20 at 0:31
  • $\begingroup$ Can you please separate your steps into different lines $\endgroup$ – rash Mar 20 at 1:19
  • $\begingroup$ @rash This way? $\endgroup$ – Rebeca Silva Mar 20 at 2:00
1
$\begingroup$

You pretty much have the answer.

$\max_{1\le i \le n}(\sum_{j=1}^n|\sum_{k = 1}^n A_{ik}.B_{kj}|)$

$\le \max_{1\le i \le n}(\sum_{j=1}^n\sum_{k = 1}^n|A_{ik}.B_{kj}|)$

$=\max_{1\le i \le n}(\sum_{j=1}^n\sum_{k = 1}^n|A_{ik}|.|B_{kj}|)$

$=\max_{1\le i \le n}(\sum_{k=1}^n(|A_{ik}|.\sum_{j=1}^n|B_{kj}|))$

$\le \max_{1\le i \le n}(\sum_{k=1}^n(|A_{ik}|. \max_{1 \le i \le n}(\sum_{j=1}^n|B_{ij}|)))$

$= \max_{1\le i \le n}(\sum_{k=1}^n|A_{ik}|) \max_{1\le i \le n}(\sum_{j=1}^n|B_{ij}|)$

$= \|A\|_\infty \|B\|_\infty.$

$\endgroup$
0
$\begingroup$

Now that you have the last inequality with the term $\sum_{j=1}^n |B_{kj}|$. Using the definition of $||\cdot||_{\infty}$ of a matrix, this term is lower than $||B||_{\infty}$ and you're done :)

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.