Machine Learning for Pest Detection and Crop Protection Optimization
Abstract
Intelligent solutions for early identification and control are essential in the face of the growing danger of pest and disease infestations in agriculture. This work presents a prediction model that uses machine learning to analyse environmental factors like humidity and temperature in order to help farmers spot possible insect assaults on chilli crops. To categorise and predict the incidence of pests using data acquired by sensors, the system employs supervised learning techniques such as Random Forest, AdaBoost, K-Nearest Neighbour (KNN), and Logistic Regression. Accuracy, Mean Squared Error (MSE), Mean Absolute Error (MAE), and Root Mean Squared Error (RMSE) were the primary performance indicators used in the extensive assessments. Random Forest outperformed the other models we looked at, with an accuracy of 85% and much lower error rates (MSE: 0.4, MAE: 0.16, RMSE: 0.63) than any of the others. AdaBoost came in second. The findings provide a solid and trustworthy decision-support tool for farmers, highlighting the efficacy of ensemble learning methods in agricultural prediction tasks. By providing data-driven insights, this solution helps with proactive pest control and supports sustainable farming practices.