Forecasting the level of expert knowledge using the GMDH method

Abstract

The expert knowledge is a very important resource for the development of an enterprise, but due to its unpredictable nature, it also very difficult to manage. Due to the growing phenomenon of frequent employee turnover in companies, it is necessary to keep such knowledge in the company and then to forecast its level in order to determine the type of missing knowledge in the company for the implementation of further projects/orders. In this paper a model for forecasting the level of expert knowledge, which assumes the use of the GMDH (Group Method of Data Handling) has been proposed. The model consists of the following elements: (1) formalised, expert knowledge acquired which is stored in the knowledge base, (2) level of knowledge in the enterprise, as the result of the clustering of acquired expert knowledge using the Bayesian network, (3) the GMDH method. Finally this model is implemented in a real case study from the Research and development department of a manufacturing company.