##
FOREWORD
The monograph "Multiparametric Statistics" by V.Serdobolskii
presents the mathematical theory
of statistical methods different by treating models
defined by a number of parameters comparable in magnitude
with sample size and exceeding it.
This branch of statistical science was developed in latter decades and
until lately was
presented only in a series of papers in mathematical
journals remaining practically unknown to the majority
of statisticians.

The first attempt to sum up these investigations
has been made in the previous author's monograph
"Multivariate Statistical Analysis.
A High-Dimensional
Approach", Kluwer Academic Publishers, 2000.

The monograph "Multiparametric Statistics" is different by more attention to
fundamentals of statistics, by refining theorems proved in the
previous book, by solving a number of new multiparametric
problems, and by the solution of infinitely-dimensional
problems.

This book is written, first, for specialists in
mathematical statistics who will find there new settings,
new mathematical methods and results of a new kind
urgently required in applications.
Actually, a new region of statistics is originated
with new mathematical problems.

Specialists in applied statistics creating
statistical packages for statistical software will be interested
in implementing new more efficient methods proposed in the book.
Advantages of these methods are obvious:
the user is liberated from the permanent
uncertainty of possible degeneration of linear methods
and gets approximately unimprovable algorithms whose quality
does not depend on distributions.

Specialists applying statistical methods
to their concrete problems
will find in the book a number of always stable
approximately unimprovable algorithms that will help them
to solve better their scientific or economic problems.

Students and postgraduates may be interested in this
book in order to get at the foremost frontier of modern statistical
science that would guarantee them the success in their future carrier.