Developing a Comprehensive Hybrid Model Utilizing Convolutional Neural Networks and Random Forest for the Advanced Classification of Tomato Rot Disease Severity Stages

This study proposed a relatively new hybrid model integrating Convolutional Neural Networks (CNN) and Random Forest (RF) for the classification of tomato fruit rot diseases into five different levels, according to graded severity levels. Breaking with the norm of the primitive and one-sided machine learning techniques in the strong categorization and the precise diagnosis of plant diseases, this research stands out for applying the deep learning potential of convolutional neural networks and the multifaceted strength of random forest for the extraction of features and the classification accordingly. The newly proposed mixed methodology, including data acquisition, preprocessing, training, and validation, indicates the hybrid model with excellent performance where the accuracy measurement was 93.5%. Besides, this portrays a significant change because it is already better compared to the current models, and it evidences its potential in the fight against weather variability and the improvement of disease management. The study results suggest that the integration of deep and ensemble learning methods in the agricultural sector is possible, wherefore their perspectives can become a base for future practical applications as a worthwhile goal. The work is substantial and represents a significant contribution to the field of agricultural informatics and plant pathology in that there is a robust tool for early disease detection and severity classification that will directly lead to quick and informative management decisions.