5 problems

The covariance kernel is … The estimators of its five largest eigenvalues are compared with the values reported by the authors of Hwang and Neumann for $3\leq n\leq 15$; for…

Dimension-free bias-reduction conjecture. Similar results hold for the estimator $f_k(\hat\Sigma)$ in the general dimension-free case.

Let $\lambda_j$ and $g_j$ denote the population eigenvalues and the relevant eigenvalue gap, and let empirical eigenvalues and eigenprojections be obtained from the empirical covar…

Let $\mathcal H$ be the underlying Hilbert space, and let $\Sigma_1,\dots,\Sigma_n$ be covariance operators on $\mathcal H$. Assume that $\Sigma_1$ is injective, and let…

Let $\Sigma_1,\dots,\Sigma_N$ be covariance operators and let $\mathscr L^0=n^{-1}\sum_{i=1}^n\Sigma_i^{1/2}$ be the initial point for the Procrustes algorithm, which iteratively a…