Development of Automated Pest Classification System Using Machine Learning Techniques for Smart Agriculture System

This research aims to develop an automated pest classification system using machine learning techniques. The study utilized a dataset of 6,444 images containing 17 different types of pests. Five machine learning algorithms were compared: Support Vector Machine (SVM), Neural Network, K-Nearest Neighbor, Logistic Regression, and Decision Tree. The results showed that SVM achieved the highest performance with 96.40% accuracy, demonstrating high Precision and Recall values across all classes and the largest area under the ROC curve. The evaluation was conducted using 10-fold Cross Validation. The system was further developed into an API for smart farming applications, enabling farmers to effectively identify and manage crop pests..