Comparison of Batches

Abstract

The second edition of this book widens the scope of the methods and applications of Applied Multivariate Statistical Analysis. We have introduced more up to date data sets in our examples. These give the text a higher degree of timeliness and add an even more applied flavour. Since multivariate statistical methods are heavily used in quantitative finance and risk management we have put more weight on the presentation of distributions and their densities. We discuss in detail different families of heavy tailed distributions (Laplace, Generalized Hy- perbolic). We also devoted a section on copulae, a new concept of dependency used in the financial risk management and credit scoring. In the chapter on computer intensive meth- ods we have added support vector machines, a new classification technique from statistical learning theory. We apply this method to bankruptcy and rating analysis of firms. The very important CART (Classification and Regression Tree) technique is also now inserted into this chapter. We give an application to rating of companies. The probably most important step towards readability and user friendliness of this book is that we have translated all Quantlets into the R and Matlab language. The algorithms can be downloaded from the authors’ web sites. In the preparation of this 2nd edition, we received helpful output from Anton Andriyashin, Ying Chen, Song Song and Uwe Ziegenhagen. We would like to thank them.