Focusing on high-dimensional applications, this 4th edition presents the tools and concepts used in multivariate data analysis in a style that is also accessible for non-mathematicians and practitioners. It surveys the basic principles and emphasizes both exploratory and inferential statistics; a new chapter on Variable Selection (Lasso, SCAD and Elastic Net) has also been added. All chapters include practical exercises that highlight applications in different multivariate data analysis fields: in quantitative financial studies, where the joint dynamics of assets are observed; in medicine, where recorded observations of subjects in different locations form the basis for reliable diagnoses and medication; and in quantitative marketing, where consumers’ preferences are collected in order to construct models of consumer behavior. All of these examples involve high to ultra-high dimensions and represent a number of major fields in big data analysis.

The fourth edition of this book on Applied Multivariate Statistical Analysis offers the following new features:

A new chapter on Variable Selection (Lasso, SCAD and Elastic Net)

All exercises are supplemented by R and MATLAB code that can be found on www.quantlet.de.

The practical exercises include solutions that can be found in Härdle, W. and Hlavka, Z., Multivariate Statistics: Exercises and Solutions. Springer Verlag, Heidelberg.

About the Author

Wolfgang Karl Härdle isa Ladislaus von Bortkiewicz Professor of Statistics at the Humboldt-Universität zu Berlin and director of C.A.S.E. (Center for Applied Statistics and Economics), director of the CRC-649 (Collaborative Research Center) “Economic Risk” and director of the IRTG 1792 “High Dimensional Non-stationary Time Series”. He teaches quantitative finance and semi-parametric statistics. His research focuses on dynamic factor models, multivariate statistics in finance and computational statistics. He is an elected member of the ISI (International Statistical Institute) and advisor to the Guanghua School of Management, Peking University.

Léopold Simaris an Emeritus Professor of Statistics at Université de Louvain, Louvain-la-Neuve, Belgium. He has been teaching mathematical statistics, multivariate analysis, bootstrap methods in statistics and econometrics in several Universities in Europe. His research focuses on non-parametric and semi-parametric methods and bootstrap techniques in statistics and econometrics. He is an elected member of the ISI and the past President of the Belgian Statistical Society. He is a regular Visiting Professor at the University of Roma, La Sapienza, Roma, Italy and at the Toulouse School of Economics, Toulouse, France.