Asymptotic behavior of eigenvalues of large rank perturbations of large random matrices
By: Ievgenii Afanasiev, Leonid Berlyand, Mariia Kiyashko
Potential Business Impact:
Makes smart computer programs learn better.
The paper is concerned with deformed Wigner random matrices. These matrices are closely connected with Deep Neural Networks (DNNs): weight matrices of trained DNNs could be represented in the form $R + S$, where $R$ is random and $S$ is highly correlated. The spectrum of such matrices plays a key role in rigorous underpinning of the novel pruning technique based on Random Matrix Theory. Mathematics has been done only for finite-rank matrix $S$. However, in practice rank may grow. In this paper we develop asymptotic analysis for the case of growing rank.
Similar Papers
On the Stability of the Jacobian Matrix in Deep Neural Networks
Machine Learning (CS)
Makes smart computer programs learn better.
On eigenvalues of a renormalized sample correlation matrix
Statistics Theory
Finds if data is related, even with lots of info.
Matrix nearness problems and eigenvalue optimization
Numerical Analysis
Makes computer math problems easier to solve.