If $A$ is chosen uniformly at random over all possible $m \times n$ (0,1)-matrices, what is the expected size of the absolute value of the determinant of $AA^T$. We can assume $m < n$ and all arithmetic is over the reals.

In Expected determinant of a random NxN matrix the expected size of the determinant for random (0,1) matrices was given as $\sqrt{(n+1)!}/ 2^n$ by David Speyer.

As pointed out by Richard Stanley, my question is incorrect in its reference to the related problem solved by David Speyer (and others). I should have said that when $A$ is a random $n$ by $n$ (0,1)-matrix then $\sqrt{E((\det A)^2)} = \sqrt{(n+1)!}/ 2^n$


By the Cauchy-Binet theorem, $\det AA^T=\sum (\det B)^2$, where $B$ ranges over all $m\times m$ submatrices of $A$. The expected value of $(\det B)^2$ is $(m+1)!/4^m$, so the expected value of $\det AA^T$ is ${n\choose m}(m+1)!/4^m$. This is also the expected value of $|\det AA^T|$ since $\det AA^T\geq 0$.

Incidentally, it is misleading to say that the expected size of the determinant of a random $n\times n$ $(0,1)$-matrix $A$ is $\sqrt{(n+1)!}/2^n$. It is true that $(n+1)!/4^n$ is the expected value of $(\det A)^2$, but this doesn't mean that $\sqrt{(n+1)!}/2^n$ is the expected value of $|\det A|$.

  • $\begingroup$ Thank you very much and thank you for the correction! I am sorry my question was a little sloppy. $\endgroup$ – Lembik Jul 2 '15 at 13:49

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.