Optimization of advertising budget allocation in the agro-industrial complex based on a machine learning algorithm

Abstract

The study is aimed at using the NSGA-III algorithm for multi-criteria optimization based on machine learning to solve marketing and advertising problems in the agro-industrial complex (AIC). In the context of increasing competitiveness and digitalization of the economy of
the Republic of Kazakhstan, the relevance of the work is due to the need to optimize the distribution of marketing budgets of enterprises, taking into account such criteria as the effectiveness of advertising campaigns, coverage of the target audience and cost reduction. The methodology used in the work involves formalization of objective functions, constraints and the use of NSGA-III to find Pareto-optimal solutions. The results obtained during cybernetic modeling usually indicate high quality of solutions (hyper-volumetric), a compromise between the criteria and the dominant role of digital channels (up to 15% increase in the budget). The data obtained during the study confirm the potential use of NSGA-III for marketing planning tasks in the agro industrial complex and can serve as a basis for making management decisions at agro-industrial enterprises. The results of the study prove that the NSGA-III algorithm has significant potential in improving marketing planning processes in the agro-industrial complex and supporting management decisions of enterprises. This approach allows for efficient use of resources in the agribusiness sector and increased competitiveness in the market.