A Comprehensive Review of Machine Learning and Deep Learning Approaches for Crop Yield Prediction
TL;DR
The analysis indicates that deep learning models generally achieve higher prediction accuracy with large datasets, while machine learning techniques remain effective for structured data.
Abstract
Crop yield prediction is essential for improving agricultural productivity, resource management, and food security in the face of climate change and increasing population demands [1], [5]. Recent advances in Machine Learning (ML) and Deep Learning (DL) have significantly enhanced the accuracy of crop yield forecasting by analyzing complex agricultural data [1], [3]. This paper presents a comprehensive review of ML- and DL-based approaches for crop yield prediction. It examines widely used algorithms such as Random Forest (RF), Support Vector Machine (SVM), Decision Tree (DT), Artificial Neural Network (ANN), XGBoost, Convolutional Neural Networks (CNN), and Long Short-Term Memory (LSTM) models [2], [3], [9]. The review also highlights the contributions of remote sensing, Unmanned Aerial Vehicles (UAVs), and Internet of Things (IoT) technologies in precision agriculture [4], [7], [11]. The analysis indicates that deep learning models generally achieve higher prediction accuracy with large datasets, while machine learning techniques remain effective for structured data. However, challenges such as limited datasets, data heterogeneity, computational complexity, and model generalization still exist [5], [6]. This review identifies current research gaps and discusses future directions for developing robust and sustainable crop yield prediction systems for smart agriculture.