Some Contributions to Multivariate Nonparametric Methods.
Abstract
Invariance principles for empirical processes, functionals of empirical distributions as well as of rank statistics provide useful tools for sophisticated studies of the distribution theory of these statistics. These results are also useful in the developing area of nonparametric sequential analysis. A general account of these invariance principles (with especial emphasis on the contributions by the principal investigator) is given here and their roles in (multivariate) nonparametric methods is discussed. Nonparametric classification procedures, mainly, the ones developed by S. K. Chatterjee, are also considered here. Finally, for the classical multivariate analysis of variance problems, especially, in the context of the so called growth curve models, the robustness of parametric procedures in the Behrens-Fisher situation, as has been studied by S. R. Chakravorti, is presented.
Document Details
- Document Type
- Technical Report
- Publication Date
- Nov 01, 1974
- Accession Number
- ADA006154
Entities
People
- P. K. Sen
Organizations
- University of North Carolina at Chapel Hill