Donald Richards
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Donald Richards has a total of 137 co-authors. Here are the top co-authors by number of shared publications:
- Frédéric Ouimet
Département de Mathématiques et D’informatique, Université du Québec à Trois-Rivières, Trois-Rivières, Canada
6 shared publications - Caroline Uhler
Massachusetts Institute of Technology, USA
5 shared publications - Christian Genest
Department of Mathematics and Statistics McGill University Montréal Québec Canada
4 shared publications Elena Hadjicosta
Pennsylvania State University
4 shared publications- Mercedes T. Richards
Department of Astronomy & Astrophysics Pennsylvania State University 16802 University Park PA U.S.A
4 shared publications - Satoshi Kuriki
The Institute of Statistical Mathematics
4 shared publications
Latest Publications
- Open Access
Stein's method for the matrix normal distribution
This work presents the first systematic development of Stein's method for matrix distributions. We establish the basic essential ingredients of Stein's method for matrix normal approximation: we derive a generator-based Stein identity from a matrix Ornstein--Uhlenbeck diffusion with two-sided scales, provide an explicit semigroup representation for the solution of the Stein equation, and obtain regularity estimates for the solution. The new methodology is illustrated with three statistical applications, these being smooth Wasserstein distance bounds to quantify the matrix central limit theorem, a Wasserstein distance bound for the matrix normal approximation of the centered matrix $T$ distribution, and the derivation of Stein's method-of-moments estimators for scale parameters of the matrix normal distribution.Open AccessStein's method for the matrix normal distribution
This work presents the first systematic development of Stein's method for matrix distributions. We establish the basic essential ingredients of Stein's method for matrix normal approximation: we derive a generator-based Stein identity from a matrix Ornstein--Uhlenbeck diffusion with two-sided scales, provide an explicit semigroup representation for the solution of the Stein equation, and obtain regularity estimates for the solution. The new methodology is illustrated with three statistical applications, these being smooth Wasserstein distance bounds to quantify the matrix central limit theorem, a Wasserstein distance bound for the matrix normal approximation of the centered matrix $T$ distribution, and the derivation of Stein's method-of-moments estimators for scale parameters of the matrix normal distribution. - Open Access
Wishart kernel density estimation for strongly mixing time series on the cone of positive definite matrices
A Wishart kernel density estimator (KDE) is introduced for density estimation in the cone of positive definite matrices. The estimator is boundary-aware and mitigates the boundary bias suffered by conventional KDEs, while remaining simple to implement. Its mean squared error, uniform strong consistency on expanding compact sets, and asymptotic normality are established under the Lebesgue measure and suitable mixing conditions. This work represents the first study of density estimation on this space under any metric. For independent observations, an asymptotic upper bound on the mean absolute error is also derived. A simulation study compares the performance of the Wishart KDE to another boundary-aware KDE that relies on the matrix-variate lognormal distribution proposed by Schwartzman [Int. Stat. Rev., 2016, 84(3), 456-486]. Results suggest that the Wishart KDE is superior for a selection of autoregressive coefficient matrices and innovation covariance matrices when estimating the stationary marginal density of a Wishart autoregressive process. To illustrate the practical utility of the Wishart KDE, an application to finance is made by estimating the marginal density function of a time series of realized covariance matrices, calculated from 5-minute intra-day returns, between the share prices of Amazon Corp. and the Standard & Poor's 500 exchange-traded fund over a one-year period. All code is publicly available via the R package ksm to facilitate implementation of the method and reproducibility of the findings.
Open AccessWishart kernel density estimation for strongly mixing time series on the cone of positive definite matrices
A Wishart kernel density estimator (KDE) is introduced for density estimation in the cone of positive definite matrices. The estimator is boundary-aware and mitigates the boundary bias suffered by conventional KDEs, while remaining simple to implement. Its mean squared error, uniform strong consistency on expanding compact sets, and asymptotic normality are established under the Lebesgue measure and suitable mixing conditions. This work represents the first study of density estimation on this space under any metric. For independent observations, an asymptotic upper bound on the mean absolute error is also derived. A simulation study compares the performance of the Wishart KDE to another boundary-aware KDE that relies on the matrix-variate lognormal distribution proposed by Schwartzman [Int. Stat. Rev., 2016, 84(3), 456-486]. Results suggest that the Wishart KDE is superior for a selection of autoregressive coefficient matrices and innovation covariance matrices when estimating the stationary marginal density of a Wishart autoregressive process. To illustrate the practical utility of the Wishart KDE, an application to finance is made by estimating the marginal density function of a time series of realized covariance matrices, calculated from 5-minute intra-day returns, between the share prices of Amazon Corp. and the Standard & Poor's 500 exchange-traded fund over a one-year period. All code is publicly available via the R package ksm to facilitate implementation of the method and reproducibility of the findings.
- 8 January 2025
- Published by Springer Nature in Statistical Papers
- Vol. 66 (1) , 1-34
EM estimation of the B-spline copula with penalized pseudo-likelihood functions
- 8 January 2025
- Published by Springer Nature in Statistical Papers
- Vol. 66 (1) , 1-34
EM estimation of the B-spline copula with penalized pseudo-likelihood functions
- 12 December 2024
- Published by American Mathematical Society (AMS) in Proceedings of the American Mathematical Society
- Vol. 153 (3) , 1299-1311
An explicit Wishart moment formula for the product of two disjoint principal minors
This paper provides the first explicit formula for the expectation of the product of two disjoint principal minors of a Wishart random matrix, solving a part of a broader problem put forth by Samuel S. Wilks [Ann. of Math. (2) 35 (1934), pp. 312–340]. The proof makes crucial use of hypergeometric functions of matrix argument and their Laplace transforms. Additionally, a Wishart generalization of the Gaussian product inequality conjecture is formulated and a stronger quantitative version is proved to hold in the case of two minors.
- 12 December 2024
- Published by American Mathematical Society (AMS) in Proceedings of the American Mathematical Society
- Vol. 153 (3) , 1299-1311
An explicit Wishart moment formula for the product of two disjoint principal minors
This paper provides the first explicit formula for the expectation of the product of two disjoint principal minors of a Wishart random matrix, solving a part of a broader problem put forth by Samuel S. Wilks [Ann. of Math. (2) 35 (1934), pp. 312–340]. The proof makes crucial use of hypergeometric functions of matrix argument and their Laplace transforms. Additionally, a Wishart generalization of the Gaussian product inequality conjecture is formulated and a stronger quantitative version is proved to hold in the case of two minors.
- Open Access
An explicit Wishart moment formula for the product of two disjoint principal minors
This paper provides the first explicit formula for the expectation of the product of two disjoint principal minors of a Wishart random matrix, solving a part of a broader problem put forth by Samuel S. Wilks in 1934 in the Annals of Mathematics. The proof makes crucial use of hypergeometric functions of matrix argument and their Laplace transforms. Additionally, a Wishart generalization of the Gaussian product inequality conjecture is formulated and a stronger quantitative version is proved to hold in the case of two minors.
Open AccessAn explicit Wishart moment formula for the product of two disjoint principal minors
This paper provides the first explicit formula for the expectation of the product of two disjoint principal minors of a Wishart random matrix, solving a part of a broader problem put forth by Samuel S. Wilks in 1934 in the Annals of Mathematics. The proof makes crucial use of hypergeometric functions of matrix argument and their Laplace transforms. Additionally, a Wishart generalization of the Gaussian product inequality conjecture is formulated and a stronger quantitative version is proved to hold in the case of two minors.
- Open Access
On Wilks' joint moment formulas for embedded principal minors of Wishart random matrices
In 1934, the American statistician Samuel S. Wilks derived remarkable formulas for the joint moments of embedded principal minors of sample covariance matrices in multivariate Gaussian populations, and he used them to compute the moments of sample statistics in various applications related to multivariate linear regression. These important but little‐known moment results were extended in 1963 by the Australian statistician A. Graham Constantine using Bartlett's decomposition. In this note, a new proof of Wilks' results is derived using the concept of iterated Schur complements, thereby bypassing Bartlett's decomposition. Furthermore, Wilks' open problem of evaluating joint moments of disjoint principal minors of Wishart random matrices is related to the Gaussian product inequality conjecture.
Open AccessOn Wilks' joint moment formulas for embedded principal minors of Wishart random matrices
In 1934, the American statistician Samuel S. Wilks derived remarkable formulas for the joint moments of embedded principal minors of sample covariance matrices in multivariate Gaussian populations, and he used them to compute the moments of sample statistics in various applications related to multivariate linear regression. These important but little‐known moment results were extended in 1963 by the Australian statistician A. Graham Constantine using Bartlett's decomposition. In this note, a new proof of Wilks' results is derived using the concept of iterated Schur complements, thereby bypassing Bartlett's decomposition. Furthermore, Wilks' open problem of evaluating joint moments of disjoint principal minors of Wishart random matrices is related to the Gaussian product inequality conjecture.
- Open Access
- 1 January 2024
- Published by Institute of Mathematical Statistics in Electronic Journal of Probability
- Vol. 29 (none) , 1-26
On the Gaussian product inequality conjecture for disjoint principal minors of Wishart random matrices
This paper extends various results related to the Gaussian product inequality (GPI) conjecture to the setting of disjoint principal minors of Wishart random matrices. This includes product-type inequalities for matrix-variate analogs of completely monotone functions and Bernstein functions of Wishart disjoint principal minors, respectively. In particular, the product-type inequalities apply to inverse determinant powers. Quantitative versions of the inequalities are also obtained when there is a mix of positive and negative exponents. Furthermore, an extended form of the GPI is shown to hold for the eigenvalues of Wishart random matrices by virtue of their law being multivariate totally positive of order 2 (MTP2). A new, unexplored avenue of research is presented to study the GPI from the point of view of elliptical distributions.
- 1 January 2024
- Published by Institute of Mathematical Statistics in Electronic Journal of Probability
- Vol. 29 (none) , 1-26
Open AccessOn the Gaussian product inequality conjecture for disjoint principal minors of Wishart random matrices
This paper extends various results related to the Gaussian product inequality (GPI) conjecture to the setting of disjoint principal minors of Wishart random matrices. This includes product-type inequalities for matrix-variate analogs of completely monotone functions and Bernstein functions of Wishart disjoint principal minors, respectively. In particular, the product-type inequalities apply to inverse determinant powers. Quantitative versions of the inequalities are also obtained when there is a mix of positive and negative exponents. Furthermore, an extended form of the GPI is shown to hold for the eigenvalues of Wishart random matrices by virtue of their law being multivariate totally positive of order 2 (MTP2). A new, unexplored avenue of research is presented to study the GPI from the point of view of elliptical distributions.
- 9 December 2023
- Published by Springer Nature in TEST
- Vol. 33 (2) , 540-563
Complete asymptotic expansions and the high-dimensional Bingham distributions
- 9 December 2023
- Published by Springer Nature in TEST
- Vol. 33 (2) , 540-563
Complete asymptotic expansions and the high-dimensional Bingham distributions
- Open Access
Complete Asymptotic Expansions and the High-Dimensional Bingham Distributions
For $d \ge 2$, let $X$ be a random vector having a Bingham distribution on $\mathcal{S}^{d-1}$, the unit sphere centered at the origin in $\R^d$, and let $Σ$ denote the symmetric matrix parameter of the distribution. Let $Ψ(Σ)$ be the normalizing constant of the distribution and let $\nabla Ψ_d(Σ)$ be the matrix of first-order partial derivatives of $Ψ(Σ)$ with respect to the entries of $Σ$. We derive complete asymptotic expansions for $Ψ(Σ)$ and $\nabla Ψ_d(Σ)$, as $d \to \infty$; these expansions are obtained subject to the growth condition that $\|Σ\|$, the Frobenius norm of $Σ$, satisfies $\|Σ\| \le γ_0 d^{r/2}$ for all $d$, where $γ_0 > 0$ and $r \in [0,1)$. Consequently, we obtain for the covariance matrix of $X$ an asymptotic expansion up to terms of arbitrary degree in $Σ$. Using a range of values of $d$ that have appeared in a variety of applications of high-dimensional spherical data analysis we tabulate the bounds on the remainder terms in the expansions of $Ψ(Σ)$ and $\nabla Ψ_d(Σ)$ and we demonstrate the rapid convergence of the bounds to zero as $r$ decreases.Open AccessComplete Asymptotic Expansions and the High-Dimensional Bingham Distributions
For $d \ge 2$, let $X$ be a random vector having a Bingham distribution on $\mathcal{S}^{d-1}$, the unit sphere centered at the origin in $\R^d$, and let $Σ$ denote the symmetric matrix parameter of the distribution. Let $Ψ(Σ)$ be the normalizing constant of the distribution and let $\nabla Ψ_d(Σ)$ be the matrix of first-order partial derivatives of $Ψ(Σ)$ with respect to the entries of $Σ$. We derive complete asymptotic expansions for $Ψ(Σ)$ and $\nabla Ψ_d(Σ)$, as $d \to \infty$; these expansions are obtained subject to the growth condition that $\|Σ\|$, the Frobenius norm of $Σ$, satisfies $\|Σ\| \le γ_0 d^{r/2}$ for all $d$, where $γ_0 > 0$ and $r \in [0,1)$. Consequently, we obtain for the covariance matrix of $X$ an asymptotic expansion up to terms of arbitrary degree in $Σ$. Using a range of values of $d$ that have appeared in a variety of applications of high-dimensional spherical data analysis we tabulate the bounds on the remainder terms in the expansions of $Ψ(Σ)$ and $\nabla Ψ_d(Σ)$ and we demonstrate the rapid convergence of the bounds to zero as $r$ decreases. - Open Access
A Reflection Formula for the Gaussian Hypergeometric Function of Matrix Argument
We obtain a reflection formula for the Gaussian hypergeometric function of real symmetric matrix argument. We also show that this result extends to the Gaussian hypergeometric function defined over the symmetric cones, and even to generalizations of the Gaussian hypergeometric function defined in terms of series of Jack polynomials. Finally, we obtain a quadratic transformation formula for the Gaussian hypergeometric function of Hermitian $2 \times 2$ matrix argument.Open AccessA Reflection Formula for the Gaussian Hypergeometric Function of Matrix Argument
We obtain a reflection formula for the Gaussian hypergeometric function of real symmetric matrix argument. We also show that this result extends to the Gaussian hypergeometric function defined over the symmetric cones, and even to generalizations of the Gaussian hypergeometric function defined in terms of series of Jack polynomials. Finally, we obtain a quadratic transformation formula for the Gaussian hypergeometric function of Hermitian $2 \times 2$ matrix argument.