The use of discriminant analysis for forecasting the limits of achievable growth of industrial enterprises.

The subject of the research is enterprises in the metallurgical industry of the Czech Republic. The choice of metallurgical enterprises in the Czech Republic as the object of research is due to the existence of a previously constructed development model of the production structure for such enterprises. The aim of the research is to forecast changes in the production structure of industrial enterprises during their growth and development. The article proposes the use of discriminant analysis to predict the limits of achievable growth for industrial enterprises within the existing production structure. Previously, a methodology for statistical forecasting of the limits of achievable growth using key predictors of organizational transformation was developed. However, unlike the statistical forecasting methodology, the use of discriminant analysis significantly reduces the labor intensity of calculations required to identify the boundaries of achievable growth of an organization within the existing production structure. Discriminant analysis was chosen as the method for conducting the research. The analysis was conducted using Wilks' method with the aid of IBM SPSS Statistics software. The choice of Wilks' method is based on its ability for algorithmic variable selection, ensuring high accuracy of the resulting model. The scientific novelty of the research lies in the scientific justification for the possibility of using discriminant analysis to forecast changes in the production structure of industrial enterprises in the process of their growth and development, based on the production structure development model. Previously, similar studies had not been conducted due to the lack of models that allow for the identification of levels of production structure development and their grouping according to these levels. The discriminant function obtained from the research results provides high accuracy in forecasting (99.7% of initial observations were classified correctly), allowing it to be used for predicting changes in the type of production structure in accordance with the development model of the production structure. The calculation of the values of the constructed discriminant function for the selected industrial enterprise allows for the classification of the enterprise (industrial enterprise, segmented enterprise, etc.) with high probability based on the previously developed production structure development model, as well as predicting directions for its further development within the existing production structure. Thus, the results of the conducted research demonstrate the potential for using the obtained statistical model to forecast changes in the production structure of industrial enterprises.