INTEGRATING BIG DATA ANALYTICS FOR PREDICTING PEST OUTBREAKS IN SMART FARMING SYSTEMS: LEVERAGING MACHINE LEARNING, CLIMATE DATA, AND REAL-TIME MONITORING

Imran Ali; Muhammad Danial Ahmad Qureshi

Authors

Imran Ali Department of Agricultural Engineering, University of Agriculture Faisalabad, Pakistan Author
Muhammad Danial Ahmad Qureshi Department of Artificial Intelligence. University of Management & Technology, Lahore, Pakistan. Author

Keywords:

Big Data, Machine Learning, Pest Outbreak Prediction, Smart Farming Systems, Climate Data, Real-Time Monitoring, Random Forest, Gradient Boosting Machines, Support Vector Machine, Artificial Neural Networks, Precision Pest Management, Sustainable Agriculture, Agricultural Data Analytics, Pest Control, IoT-based Pest Monitoring

Abstract

The agriculture world now deals with growing problems of safe farming practices and pest prevention across all countries. The normal ways to control pests create high costs and do little while also hurting our environment. By blending large data analysis tools and machine learning methods researchers studied ways to spot insect outbreaks in smart farming systems. The research team studied whether combining climatic data, past bug information with current surveillance data would enhance pest forecasting accuracy. We tested four machine learning systems to detect insect outbreaks which included random forest, support vector machine (SVM), gradient boosting machine (GBM), and artificial neural network (ANN). The GBM and random forest methods produced superior outcomes than other models although random forest demonstrated the highest prediction accuracy rate (85%). Incorporating real-time monitoring data with climatic factors enhanced forecasting results between 5% and 8%. The research reveals that analyzing big data with machine learning techniques helps farmers protect their land while reducing pesticide use and staying in charge of crop pests.