Přehled o publikaci

The paper represents a continuation of our previous results, which were closely linked to the topic of automated search for factors of corporate competitiveness and financial performance. As opposed to the results presented at ICMLG 2014 and ECMLG 2014, the current research deals with other characteristics of organizations not investigated before and related to Corporate Governance, like their legal form, the international diversity of top management (domestic only or also foreign) and particularly the non-anonymity of the ownership structure.The data were gathered in the period 2011 to 2013 from 222 companies with various legal forms. The main purpose of our research was to investigate which of these characteristics have an impact on corporate financial performance that has been assessed by the Return on Assets (ROA) index. The paper attempts to answer the research question of whether the ownership structure, particularly the fact whether the organization or enterprise owners are or are not anonymous, has a major influence on corporate financial performance. The methodology used in analysing and processing the data had to respect the fact that the characteristics (variables) of the companies which were individually investigated are not mutually independent, thus multidimensional methods have to be used. Therefore, we used here our non-linear kernel regression model which had already been successfully verified, having been developed in the field of statistical pattern recognition. Prediction error of the proposed model is then used as a feature selection criterion in the process of identifying factors that affect Corporate Governance and the financial performance the most. The results presented in the paper demonstrate that the type of ownership structure (anonymous or non-anonymous) has a dominant influence on the financial performance among the investigated characteristics. The target audience includes researchers in the fields of management and business science.