Testing Compound Symmetry in a Normal Multivariate Distribution

Testing Group Symmetry of a Multivariate Distribution

We propose and study a general class of tests for group symmetry of a multivariate distribution, which encompasses different types of symmetry, such as ellipsoidal and permutation symmetries among others. Our approach is based on supremum norms of special empirical processes combined with bootstrap. We show that these tests are consistent against any fixed alternative. This work generalizes the...

Screening for a Multivariate Mixture Normal Distribution

Multi-level multivariate normal distribution with self similar compound symmetry covariance matrix

We study multi-level multivariate normal distribution with self similar compound symmetry covariance structure for k different levels of the multivariate data. Both maximum likelihood and unbiased estimates of the matrix parameters are obtained. The spectral decomposition of the new covariance structure are discussed and are demonstrated with a real dataset from medical studies.

HYPOTHESIS TESTING FOR AN EXCHANGEABLE NORMAL DISTRIBUTION

Consider an exchangeable normal vector with parameters ????2, and ?. On the basis of a vector observation some tests about these parameters are found and their properties are discussed. A simulation study for these tests and a few nonparametric tests are presented. Some advantages and disadvantages of these tests are discussed and a few applications are given.

screening for a multivariate mixture normal distribution