Take the 2-minute tour ×
Mathematics Stack Exchange is a question and answer site for people studying math at any level and professionals in related fields. It's 100% free, no registration required.

Suppose we have two matrices, A and B, and

$\left\Vert A\right\Vert _{F}\geq\left\Vert B\right\Vert _{F}$

where $\left\Vert .\right\Vert _{F}$ denotes Frobenius norm. Does it imply

$\left\Vert A\right\Vert \geq\left\Vert B\right\Vert $

for any other norm $\left\Vert .\right\Vert $ ??

Thanks

share|improve this question
    
Yes but with a multiplicative constant in the inequality, since all finite dimensional norms are equivalent. –  soup Dec 23 '12 at 10:58

1 Answer 1

up vote 0 down vote accepted

No. It is not the case. For instance, let $A,B$ be $2 \times 2$ matrices. Let the singular values of $A$ be $\sigma_{1A}, \sigma_{2A}$ and that of $B$ be $\sigma_{1B}, \sigma_{2B}$. $$\Vert A \Vert_F \geq \Vert B \Vert_F \implies \sigma_{1A}^2 + \sigma_{2A}^2 \geq \sigma_{1B}^2 + \sigma_{2B}^2$$ This doesn't mean that $$\Vert A \Vert_2 = \sigma_{1A} \geq \sigma_{1B} = \Vert B \Vert_2$$ For instance, let $\sigma_{1A} = \sigma_{2A} = 1$ and $\sigma_{1B} = 1.5, \sigma_{2B} = 0$.

We then have $$\Vert A \Vert_F = \sqrt{2} > \sqrt{1.5} = \Vert B \Vert_F$$ but $$\Vert A \Vert_2 = 1 < \sqrt{1.5} = \Vert B \Vert_2$$

share|improve this answer
    
thanks Marvis :) –  aning Dec 23 '12 at 12:26

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.

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