MODELING OF THE SYNTHETIC INDICATOR OF COMPETITIVENESS OF AGRICULTURAL ENTERPRISES: A METHODOLOGICAL APPROACH TO THE USE OF NEURAL NETWORK TOOLS

Abstract

The article is devoted to the development of a methodical approach to modelling a synthetic indicator of the competitiveness of agricultural enterprises using the tools of neural networks. The authors used general scientific and special research methods, such as monographic, logical-theoretical, statistical and economic-mathematical, visualization, system analysis, taxonomy and neural network modelling, generalization, logical abstraction and conclusion generation. The study was based on materials from the State Statistics Service of Ukraine, scientific developments of foreign and domestic scientists on the defined topic, and financial statements of the agricultural enterprises of Vinnytsia region LLC «Ahrokompleks «Zelena dolyna», PJSC «Dashkivtsi», LLC «Selyshchanske», PE «Dary sadiv», PE «Fortuna» the main type of economic activity of which according to Classification of economic activities 01.11 – cultivation of cereals (except rice), legumes and oilseeds. The article develops and presents a non-classical approach to the assessment of the competitiveness of agricultural enterprises has been developed, which is based on the principles of neural network modelling. It allows to obtain a well-founded quantitative indicator, which can be easily interpreted into a linguistic evaluation on a three-level scale of competitiveness and used for comparison, monitoring and making sound decisions on improving the competitiveness of agricultural enterprises. The non-classical approach complements traditional methods of competitiveness assessment, expanding their capabilities and eliminating certain limitations. The use of neural network modelling in competitiveness assessment allows to take into account complex and non-linear relationships between different factors and indicators, which contributes to an increase in the objectivity and accuracy of competitiveness assessment, which in turn allows enterprises to make better decisions and improve their strategies to achieve success in the market. The results of the study can be used to support strategic decision-making in the agricultural sector, identify priority development directions, and improve the competitive strategies of enterprises and the functioning of business processes.

Keywords

• agricultural enterprises
• synthetic index
• competitiveness
• neural networks
• latent trait
• modeling