Abstract

In agriculture, optimizing crop yield and maintenance practices is essential for ensuring food security and sustainable farming. Traditional approaches often lack the efficiency needed to process large agricultural datasets and accurately predict yield under varying environmental conditions. This project leverages the Light Gradient Boosting Machine (LightGBM), a high-performance, gradient- boosting framework specifically designed for large-scale data handling, to address the challenge of yield prediction and crop maintenance optimization. By integrating LightGBM, which handles heterogeneous data with high accuracy, we aim to enhance predictions on crop yield while minimizing resource use. The proposed method analyzes a range of factors, including soil quality, weather conditions, irrigation practices, and historical crop yield records. Initial results indicate that LightGBM outperforms conventional models with a 94.7% accuracy rate in yield prediction and reduces maintenance costs by up to 20% by recommending optimized agricultural practices based on specific environmental conditions. These findings underscore the potential of LightGBM as an effective tool in precision agriculture, ultimately aiding farmers in making informed decisions and improving agricultural productivity.

Authors

Sunil Kumar¹, Mohammed Ali Sohail², Sandhya Jadhav³, Raj Kumar Gupta⁴
AURO University, India¹, Jazan University, Kingdom of Saudi Arabia², Bharati Vidyapeeth College of Engineering, India³, Sardar Vallabhbhai Patel College, India⁴

Keywords

Precision Agriculture, Yield Prediction, LightGBM, Crop Maintenance, Data Optimization