Machine Learning approaches for Crop Yield Prediction: A Review

Abstract

By providing food producers with much greater access to data about their operations, artificial intelligence (AI) in agriculture has revolutionized the way that agricultural operations throughout the globe operate. A farming method known as yield mapping employs supervised machine learning algorithms to find patterns in massive data sets that may be utilized for crop planning. A critical problem in agriculture is estimating increased crop output with machine learning algorithms. The current study presents a detailed analysis of the applications of ML for predicting crop yield for different datasets. The research papers have been selected for this review, focused on the latest publications, which suggests how vital is this research field. The techniques, traits, and qualities employed in studies related to crop yield prediction were extracted from that research and correlated in this study. In the models, soil type, rainfall, season, and temperature are the most often utilized characteristics, and Random Forest is the most frequently used method, according to our research. Deep learning-based studies were also analyzed, and the most commonly used Deep Learning technique is DNN and CNN.