Analysis of trading companies on the basis of Kohonen networks

The contribution deals with the analysis of trade companies using artificial neural networks, more specifically Kohonen networks, where the basis for the analysis is primarily the data from evaluating economic performance of a business entity, where each entity strives for increasing its market valu...

Full description

Saved in:
Bibliographic Details
Published inLittera scripta (Online) no. 1/2019; pp. 62 - 76
Main Author Tomáš Krulický, Zuzana Rowland
Format Journal Article
LanguageEnglish
Published Institute of Technology and Business, České Budějovice 01.08.2019
Subjects
cluster analysis
economic performance of company
Kohonen networks
neural networks
trade
Online AccessGet full text

Cover

More Information
Summary:The contribution deals with the analysis of trade companies using artificial neural networks, more specifically Kohonen networks, where the basis for the analysis is primarily the data from evaluating economic performance of a business entity, where each entity strives for increasing its market value. This contribution also estimates the development of trade companies. For the analysis, the data of 11,604 companies in the given sector in the Czech Republic were used. Cluster analysis is carried out using neural networks. From the clusters, those with the highest number of companies were chosen. Subsequently, the analysis of absolute values of selected financial statements items is carried out. From the owners´ point of view, the situation is not positive, as the rate of deposits appreciation is even lower than the risk-free appreciation of 10-year state bonds. In the conclusion, recommendations for trade companies are given.
ISSN:1805-9112