A. K. Gupta & V. L. Girko 
Multidimensional Statistical Analysis and Theory of Random Matrices [PDF ebook] 
Proceedings of the Sixth Eugene Lukacs Symposium, Bowling Green, Ohio, USA, 29–30 March 1996

Frontmatter — PREFACE — Welcoming remarks — Principal Procedures for Statistical Analysis of Random Arrays (SARA) — CONTENTS — Some remarks and a bibliography on the Kantorovich inequality — Band random matrices and quantum chaos — Weighted continuous metric scaling — Canonical equation for the resolvent of empirical covariance matrices pencil — Strong Law for the eigenvalues of empirical covariance matrices — On the asymptotic behavior of Markov chains with small random perturbations of transition probabilities — Multivariate statistical inference involving circulant matrices: A review — Asymptotics of eigenvalue-normed eigenvectors of sample variance and correlation matrices — Formal density expansions via multivariate mixtures — Double shrinkage estimators of ratio of variances — On sequential search for significant variables of unknown function — Heirarchical random matrices and operators. Application to Anderson model — Asymptotic properties of invariant tests — New classes of probability inequalities for some classical distributions — Equivariant estimation of a subspace — Optimal design of experiments for multivariate response in two-factor linear models — Multivariate analysis: Does it really work in statistical applications? — Robust estimation and testing the mean vector — Detection of outliers in the presence of multicollinearity — Estimation in multivariate elliptically contoured linear models II — Mean and covariance structure analysis with missing data — Multivariate elliptically contoured linear models and some aspects of the theory of random matrices