Hybrid Adaptive Method for Optimizing Marketing Strategies for the Agro-industrial Complex Based on the NSGA-III Algorithm and Machine Learning Methods

The article presents a hybrid adaptive method for optimizing marketing strategies of the agro-industrial complex of Kazakhstan (agro-industrial complex) based on the modified NSGA-III algorithm and machine learning (ML) methods. The proposed method is aimed at finding Pareto – optimal solutions that take into account three main criteria-the efficiency of distribution channels. f1(ꭓ), audience access f2(ꭓ) and costs f3(ꭓ). The problem of multi-criterion optimization has been formalized with restrictions that reflect the specifics of the agro-industrial complex, such as minimum shares. Digital (≥20%) and Print (≥10%), as well as combined restrictions Digital and Events (≥30%). NSGA-III adaptation mechanisms have been implemented, including support Route generation (Das-Dennis) and evolutionary operators (SBX, PM), supplemented by KMeans clustering, for Solution Analysis. The proposed composite indicator F(ꭓ)=(0.6·f1(ꭓ)+0.4·f2(ꭓ))/(f3(ꭓ)+0.1) made it possible to determine strategies. The simulation results show the priority of TV (47.7%).