1
$\begingroup$

I have a a full column rank matrix A, and using this I want to construct a matrix with spectral radius less than 1. I do that using,

H = $I-\alpha A^{T} A$ ($I$ is identity matrix), where the term $\alpha$ is defined as, $0<\alpha < \frac{2}{trace(A^{T} A)}$.

Can I claim that each one of therm in my matrix H will be less than or equal to 1?

I believe that the solution is yes, since I have not been able to come up with the a single scenario where this claim is not true. But I am not able to prove this.

$\endgroup$
2
  • $\begingroup$ No I mean how can I prove that the values in H are bounded by 1? $\endgroup$ Mar 29, 2016 at 2:50
  • $\begingroup$ Sorry, I misread the question. I have deleted my comment. $\endgroup$
    – LSpice
    Mar 29, 2016 at 2:51

1 Answer 1

3
$\begingroup$

Yes. Your matrix $H$ is symmetric, so the spectral radius is the same as the maximal eigenvalue is the same as the maximal singular value. From the singular value description, for any unit vectors $\vec{u}$ and $\vec{v}$, we deduce that $|\vec{u}^T H \vec{v}| < 1$. Taking $\vec{u}$ and $\vec{v}$ to be the $i$-th and $j$-th basis vector, we have $|H_{ij}| < 1$.

$\endgroup$
2
  • 2
    $\begingroup$ To say it another way, for a symmetric matrix, the spectral radius equals the operator norm. And by Cauchy-Schwarz $|u^T H v| \le |u| |Hv| \le |u| |v| \|H\|$. Then take $u,v$ to be unit basis vectors. $\endgroup$ Mar 29, 2016 at 4:10
  • 2
    $\begingroup$ @RohitShukla : but when the matrix is not symmetric, it is not true anymore, take $M = \left(\begin{array}{ll}1&2000000\\0&0\end{array}\right)$ whose eigenvalues are $1$ and $0$, hence its spectral radius is $1$ ... $\endgroup$
    – reuns
    Mar 29, 2016 at 8:11

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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