Marchenko–Pastur distribution

In the mathematical theory of random matrices, the Marchenko–Pastur distribution, or Marchenko–Pastur law, describes the asymptotic behavior of singular values of large rectangular random matrices. The theorem is named after soviet mathematicians Volodymyr Marchenko and Leonid Pastur who proved this result in 1967.

Plot of the Marchenko-Pastur distribution for various values of lambda

If denotes a random matrix whose entries are independent identically distributed random variables with mean 0 and variance , let

and let be the eigenvalues of (viewed as random variables). Finally, consider the random measure

counting the number of eigenvalues in the subset included in .

Theorem. [citation needed] Assume that so that the ratio . Then (in weak* topology in distribution), where

and

with

The Marchenko–Pastur law also arises as the free Poisson law in free probability theory, having rate and jump size .

Moments

edit

For each  , its  -th moment is[1]

 

Some transforms of this law

edit

The Stieltjes transform is given by

 

for complex numbers z of positive imaginary part, where the complex square root is also taken to have positive imaginary part.[2] The Stieltjes transform can be repackaged in the form of the R-transform, which is given by[3]

 

The S-transform is given by[3]

 

For the case of  , the  -transform [3] is given by   where   satisfies the Marchenko-Pastur law.

 

where  

A convenient form is often   which simplifies to

 

Application to correlation matrices

edit

For the special case of correlation matrices, we know that   and  . This bounds the probability mass over the interval defined by

 

Since this distribution describes the spectrum of random matrices with mean 0, the eigenvalues of correlation matrices that fall inside of the aforementioned interval could be considered spurious or noise. For instance, obtaining a correlation matrix of 10 stock returns calculated over a 252 trading days period would render  . Thus, out of 10 eigenvalues of said correlation matrix, only the values higher than 1.43 would be considered significantly different from random.

See also

edit

References

edit
  1. ^ Bai & Silverstein 2010, Section 3.1.1.
  2. ^ Bai & Silverstein 2010, Section 3.3.1.
  3. ^ a b c Tulino & Verdú 2004, Section 2.2.
  • Bai, Zhidong; Silverstein, Jack W. (2010). Spectral analysis of large dimensional random matrices. Springer Series in Statistics (Second edition of 2006 original ed.). New York: Springer. doi:10.1007/978-1-4419-0661-8. ISBN 978-1-4419-0660-1. MR 2567175. Zbl 1301.60002.
  • Epps, Brenden; Krivitzky, Eric M. (2019). "Singular value decomposition of noisy data: mode corruption". Experiments in Fluids. 60 (8): 1–30. Bibcode:2019ExFl...60..121E. doi:10.1007/s00348-019-2761-y. S2CID 198436243.
  • Götze, F.; Tikhomirov, A. (2004). "Rate of convergence in probability to the Marchenko–Pastur law". Bernoulli. 10 (3): 503–548. doi:10.3150/bj/1089206408.
  • Marchenko, V. A.; Pastur, L. A. (1967). "Распределение собственных значений в некоторых ансамблях случайных матриц" [Distribution of eigenvalues for some sets of random matrices]. Mat. Sb. N.S. (in Russian). 72 (114:4): 507–536. Bibcode:1967SbMat...1..457M. doi:10.1070/SM1967v001n04ABEH001994. Link to free-access pdf of Russian version
  • Nica, A.; Speicher, R. (2006). Lectures on the Combinatorics of Free probability theory. Cambridge Univ. Press. pp. 204, 368. ISBN 0-521-85852-6. Link to free download Another free access site
  • Tulino, Antonia M.; Verdú, Sergio (2004). "Random matrix theory and wireless communications". Foundations and Trends in Communications and Information Theory. 1 (1): 1–182. doi:10.1561/0100000001. Zbl 1143.94303.
  • Zhang, W.; Abreu, G.; Inamori, M.; Sanada, Y. (2011). "Spectrum sensing algorithms via finite random matrices". IEEE Transactions on Communications. 60 (1): 164–175. doi:10.1109/TCOMM.2011.112311.100721. S2CID 206642535.