The analysis and re-optimization of food systems by using intelligent optimization algorithms and machine learning

Abstract

Facing the serious concern of global food security, this study focuses on the feasibility of improvements to food systems and predicts future changes. This study proposes a coordination evaluation index incorporating diverse indicators of different food systems, analyzes the changes after modification, and predicts the equilibrium time and the critical point. In this regard, we pre-processed the data using the entropy method, variance contribution rate, and normalization. Instead of utilizing single-element linear forecasting and economic income and expenditure models, we innovatively developed a multivariate evaluation system for population, cultivated land, and food systems in three major directions. Meanwhile, after conducting a cross-sectional comparison of the prediction effects of various algorithms, we finally selected Gaussian process regression and a neural network to build a prediction model to develop food systems of different sizes. After establishing the evaluation index and development prediction model, we fitted the three-dimensional surface of the developmental change using thin-plate interpolation. We adopted the swarm intelligence optimization algorithm to search for the balance and critical points after the change. We also compared various swarm intelligence optimization algorithms, such as the particle swarm optimization algorithm, salp swarm algorithm, and whale optimization algorithm.