4
$\begingroup$

Suppose I have $n$ data vector $X_1, ..., X_n$, where each $X_i$ is a length $p$ random vector.

The sample covariance matrix is $S=\frac{1}{n-1}\sum_i (X_i-\bar{X})(X_i-\bar{X})^T$. If I want to compute the generalized variance, I can compute the determinant of $S$, i.e. $|S|$.

In my actual use case where $p$ is fixed, but $p>n$, is $|S|$ the best estimator that I can use?

In my literature search I come across two types of papers, one that address the case when $p<n$ and talks about several different kinds of better estimators. The other talks about the case when $p$ can grow, but $p(n)<n$, and they talk about asymptotic results. This later case confuses me a bit because in my case, since $p$ is fixed, $n$ will become larger than $p$ in the asymptotics, but in practice I will never be in that case.

Edit

My purpose is to use the generalized variance as summary statistics. I know using a one number summary to describe the variability of multivariate data (functional data actually, in my case) is not a brilliant idea. But that is what is needed for practical purpose. So $\hat{\Sigma}$ (whatever estimator we use) has $p\times p$ many numbers (or $pC2$ many), which is too much. A one number summary like $tr(\hat{\Sigma})$ or $|\hat{\Sigma}|$ it better.

$\endgroup$
5
  • 1
    $\begingroup$ The answer would depend on your goal. You could use any of the eigenvalues of the covariance matrix, or a combination of them. (or equivalently, use singular values of the centered/scaled data matrix). The determinant is the product of these (but really you would use their geometric average, to be a "variance", i.e. take n-th root). $\endgroup$
    – GeoMatt22
    Commented Apr 12, 2017 at 5:44
  • $\begingroup$ @GeoMatt22 Some of the singular values are numerically equivalent to zero. What are you supposed to do in this case? $\endgroup$ Commented Apr 12, 2017 at 7:24
  • $\begingroup$ Just multiply the non-zero ones? $\endgroup$ Commented Apr 12, 2017 at 11:55
  • $\begingroup$ @MatthewGunn I was implicitly considering the reduced SVD, so essentially as in your answer (i.e. "structural" zeros eliminated, but "full rank" = $n$, svd( ,'econ') in Matlab; if there are additional numerical zeros, these should probably also be eliminated?) $\endgroup$
    – GeoMatt22
    Commented Apr 12, 2017 at 13:26
  • $\begingroup$ Thanks GeoMatt22. In practice I will be using software to compute all these. So for example in r, determinant() uses LU decomposition (if I have not misunderstood). But my question is not really on how to compute $|S|$, but whether $|S|$ is the best estimator to use. For example, some of the papers I cited use some sort of shrinkage estimator $aS+b\Phi$, or some bias corrected form like $S+\tau$. I am not saying these are better (or not), my question is whether $|S|$ is a good estimator. $\endgroup$
    – qoheleth
    Commented Apr 12, 2017 at 23:49

1 Answer 1

3
$\begingroup$

The problem with $p \geq n$

If $p \geq n$ then as you recognize your covariance matrix is necessarily rank deficient.

Consider $p=2$ case. If $n=3$ you can find the ellipse defined by the estimated covariance matrix for the three points.

enter image description here enter image description here

But if $n=2$, then what is the ellipse supposed to be?! The rank deficient covariance matrix defines the degenerate ellipse that's a line segment between the two points. For $p \geq n$, the usual sample covariance has some dimension with zero variation (in this case, basically the $(1,-1)$ vector). Does that mean there's actually zero variation on that dimension, that your data only occupies a one dimensional space? (Or if we drew a third blue point, could it fall within one of the larger ellipses?)

If you're a Bayesian, what's your prior on the covariance matrix? A direction perhaps to go is to add some kind of prior or regularization? (I'm admittedly not an expert in this area. I'd love to see other answers.)

And I think we'll have to know more about your problem and why you want the determinant to have an idea of how to proceed? Perhaps you want something like the pseudo-determinant?

$\endgroup$
0

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.