Machine Learning Models for Detection and Prediction of Crop Diseases : A Review

Abstract: This paper aims to provide a comprehensive overview of machine learning (ML) techniques across various data types, fostering opportunities to address research gaps and advance the field, particularly in the detection and prediction of crop diseases. The survey presents valuable insights into ML-based techniques for forecasting, detecting, and classifying diseases and pests. It highlights the importance of maintaining long-term datasets encompassing weather, disease, and pest data. Time-series ML models, such as recurrent neural networks (RNNs), are shown to be effective tools for accurately predicting disease and pest occurrences based on sequences of meteorological measurements. Additionally, incorporating normalized difference vegetation index (NDVI) measurements can provide supplementary insights into crop development. Leveraging computer vision and deep learning algorithms, particularly convolutional neural network (CNN) models, proves advantageous for detecting and classifying pests and diseases, outperforming traditional approaches that rely on manual feature extraction.

Keywords: Machine Learning, Support Vector Machine, Random Forest, Artificial Neural Networks, Plant disease detection.