An Effective Soil Analysis and Crop Yield Prediction Based on Optimised Light GBM in Smart Agriculture

Abstract

In the agricultural sector, crop yield prediction plays an important role as it helps farmers make decisions about the growing season and type of crops to get better yield. The main goal in the agricultural sector is to reduce operating costs and pollution by improving crop yields and quality. This paper proposes an effective method for soil analysis and crop yield prediction for intelligent agriculture. The collected data are preprocessed using missing value interpolation and data normalisation techniques. Feature selection is performed on the preprocessed data using the Aquila-based adaptive optimisation algorithm, which selects the best trait subset for yield prediction. An improved lightweight gradient-boosting machine based on the Battle Royale Optimisation technique is used for classification. The performance of the proposed system is evaluated using mean absolute error, root mean square error, R-squared, mean square error, mean square logarithmic error and mean absolute percentage error, and the proposed system achieved an accuracy of 97%. The proposed system accurately predicts crop yields, improving crop production and quality.