Selection of Informative Factors in Merger and Acquisition Process

P. Pudil, P. Pirožek, and P. Somol (Czech Republic)

Keywords

feature selection, approximation of class conditional densities, pseudo-Bayes decision rule, merger and acquisition process

Abstract

The process of merger and acquisition with respect to its expected success is analyzed by means of pat tern recognition techniques. More specifically, so called mixture model for feature selection and pattern recognition, convenient for cases with absent a priori information on underlying probability distributions, has been used. It helped to determine the most informative factors which discriminate successful mergers and acquisitions from unsuccessful ones. The data from the privatization process in the Czech Republic have been used to test the methodology.

Important Links:



Go Back