Enhanced Identification of Apple Leaf Diseases Through Optimized Machine Learning Algorithms

Abstract

Apple orchards play a vital role in global food production, but they face constant threats from diseases like scab, rust, and rot. Accurate and timely identification of these diseases is paramount to prevent crop losses and ensure food security. In this study, we present a novel approach that integrates the Firefly Optimization (FFO) algorithm with Rough Set Theory (RST) to optimize feature selection for the identification of apple leaf diseases. We employ the K - Nearest Neighbors (k - NN) machine learning algorithm to classify various types of apple leaves, including healthy, rot, rust, and scab leaves. The results reveal the remarkable effectiveness of the optimized k - NN algorithm in comparison to the non - optimized version. The optimized approach significantly improves the True Positive Rate (TPR) and reduces the False Negative Rate (FNR) for each type of apple leaf disease, indicating enhanced accuracy in disease identification. These findings have substantial practical implications for apple growers, enabling informed management decisions to reduce crop losses. Furthermore, this study underscores the potential of machine learning, image processing, and feature selection techniques in advancing disease identification accuracy, with broader applications in agriculture. Ultimately, this research contributes to sustainable agricultural practices and global food security by harnessing technology to combat crop diseases effectively.