Robust Multivariate Analysis

Robust methods are presented for multivariate analysis, using large sample theory and illustrated for methods such as principal component analysis, canonical correlation analysis, and factor analysis. A simple bootstrap confidence region method is also provided. The text is suitable for a first course in Multivariate Statistical Analysis or a first course in Robust Statistics, and is useful for people familiar with traditional multivariate topics who want to know more about handling data sets with outliers.

\n Format: Hardback
\n Length: 501 pages
\n Publication date: 13 December 2017
\n Publisher: Springer International Publishing AG
\n

Robust methods are those that are resilient to the assumption of a multivariate normal distribution or to specific types of outliers. Instead of relying on the complex theory derived from the multivariate normal distribution, the text presents a simpler and more practical large sample theory. The robust regression and robust multivariate location and dispersion estimators, supported by theoretical foundations, are developed. The robust techniques are demonstrated for methods such as principal component analysis, canonical correlation analysis, and factor analysis. Additionally, a straightforward approach to bootstrapping confidence regions is provided.

This book offers a wealth of research on robust multivariate analysis, with many of the findings being published for the first time. It serves as an excellent introduction to Multivariate Statistical Analysis or Robust Statistics, suitable for both undergraduate and graduate students. Those with a background in traditional multivariate topics will find this text particularly valuable in learning how to handle data sets with outliers.

The authors have made numerous R programs and R data sets available on their website, further enhancing the usefulness of the text for practical applications.

\n Weight: 996g\n
Dimension: 242 x 172 x 30 (mm)\n
ISBN-13: 9783319682518\n
Edition number: 1st ed. 2017\n