A review of machine learning approaches for predicting lettuce yield in hydroponic systems

Accurate and timely yield prediction of hydroponically grown lettuce is essential for financial planning, strategic decision-making, and enhancing farmers' profitability. In controlled hydroponic environments, this prediction remains challenging, mainly due to complex factors influencing growth. Machine Learning (ML) offers advanced methods to address these challenges. This review analyzes ML techniques for forecasting lettuce yield in hydroponic systems, starting with an overview of global trends in lettuce production. It then explores core ML methodologies, key model characteristics, and application-specific features that contribute to yield prediction. A comparative analysis of existing ML models also highlights their strengths and limitations. Current challenges, such as data integration and prediction accuracy, are discussed alongside potential improvements through remote sensing, monitoring, and feature optimization. This paper concludes by proposing a framework aimed at efficient yield prediction in hydroponics, offering insights for future research and applications in agricultural technology.