Page 1 Next

Displaying 1 – 20 of 91

Showing per page

A homogeneity test of large dimensional covariance matrices under non-normality

M. Rauf Ahmad (2018)


A test statistic for homogeneity of two or more covariance matrices is presented when the distributions may be non-normal and the dimension may exceed the sample size. Using the Frobenius norm of the difference of null and alternative hypotheses, the statistic is constructed as a linear combination of consistent, location-invariant, estimators of trace functions that constitute the norm. These estimators are defined as U -statistics and the corresponding theory is exploited to derive the normal limit...

Canonic inference and commutative orthogonal block structure

Francisco P. Carvalho, João Tiago Mexia, M. Manuela Oliveira (2008)

Discussiones Mathematicae Probability and Statistics

It is shown how to define the canonic formulation for orthogonal models associated to commutative Jordan algebras. This canonic formulation is then used to carry out inference. The case of models with commutative orthogonal block structures is stressed out.

Classification into two von Mises distributions with unknown mean directions

Kryštof Eben (1983)

Aplikace matematiky

The paper deals with two Mises distributions on the circle with unknown mean directions and a common concentration parameter that is known. The likelihood rule and the plug-in rule are examined. For the statistic of the plug-in rule, the moment generating function is given and a method of obtaining the moments is proposed.

Classifiers for doubly multivariate data

Mirosław Krzyśko, Michał Skorzybut, Waldemar Wołyński (2011)

Discussiones Mathematicae Probability and Statistics

This paper proposes new classifiers under the assumption of multivariate normality for multivariate repeated measures data (doubly multivariate data) with Kronecker product covariance structures. These classifiers are especially useful when the number of observations is not large enough to estimate the covariance matrices, and thus the traditional classifiers fail. The quality of these new classifiers is examined on some real data. Computational schemes for maximum likelihood estimates of required...

Contrastación de hipótesis en diseños multivariados split-plot con matrices de dispersión arbitrarias.

Guillermo Vallejo Seco, José Ramón Escudero García, Angel M. Fidalgo Aliste, M. Paula Fernández García (2000)


El presente trabajo examina diversos procedimientos para contrastar hipótesis nulas globales, correspondientes a datos obtenidos mediante diseños multivariados split-plot cuando se incumple el supuesto de homogeneidad de las matrices de dispersión. Un examen de estos procedimientos para un amplio número de variables confirma, por un lado, la robustez del procedimiento multivariado de Welch-James dado por Johansen (1980) para probar el efecto principal de los ensayos y, por otro, la robustez de la...

Currently displaying 1 – 20 of 91

Page 1 Next