ARCHIVES

Federated Machine Learning (FML) has become an increasingly common approach for addressing data security and privacy issues in fields such as the Internet of Things (IoT) and agriculture. The application of FML in crop yield prediction is the primary objective of this review paper. Decentralized data from agricultural systems and Internet of Things devices is used to train models cooperatively without compromising sensitive data. FML ensures that data remains local by enabling distributed learning across several edge devices, improving privacy, and enabling effective and precise agricultural output forecasting. IoT sensors gather a variety of data in smart agriculture, including plant health measurements, soil conditions, and weather patterns. These data may be easily included in federated learning frameworks to provide real-time insights and decision-making. This study examines recent developments in agriculture-specific federated learning approaches, emphasizing important issues including communication effectiveness, model personalization, and data heterogeneity. The study additionally examines new developments in federated learning and innovative machine learning approaches to increase prediction accuracy, lower computing costs, and offer dependable options for sustainable agriculture.