AI in food inequality: Leveraging artificial intelligence to predict food waste in agriculture and post-harvesting

Food disparity is an international trend, driven by inefficiencies in poorly developed food distribution and agricultural infrastructure. An FAO and Kaggle Datasets study estimates post-harvest losses as intervention points with global median losses at 19.8%. India, as a major producer of most food commodities in agriculture, has relatively low post-harvest losses (3.2%), yet suffers from chronic hunger, as is clear from its 111/125 ranking on the Global Hunger Index [9][13]. This paradox of high production but low consumer supply outcome emphasizes the need for a critical analysis of India. This study utilized machine learning (ML) models in the form of gradient boosting regression to analyze Indian farm data, including such variables as pesticide, fertilizer, farm size, crop type, harvest date, and climatic conditions. The optimal model had an R 2 measure of 0.999 in predicting best farming practice based on local conditions. The optimization model increased food retention after post-harvest by 3.42% over modern methods, bringing food into the supply chain at no extra cost. Lastly, these findings present actionable recommendations to future agricultural policy while also offering practical solutions to regions facing analogous food security concerns.