Crop yield prediction in diverse environmental conditions using ensemble learning

Abstract

Precise assessment of crop yield is a vital component in agricultural planning and decision-making, having immediate consequences for food security and allocation of resources. This study presents a new approach for predicting agricultural output in different climatic conditions by integrating the xgboost algorithm with the Whale Optimization Algorithm (WOA). XGBoost is a kind of ensemble learning method that enhances the accuracy of predictions by combining the results of many weak learners. However, the performance of the system may be significantly affected by the selection of suitable hyper parameters and feature subsets. To address this problem, we use the WOA algorithm, a nature-inspired optimization approach that mimics the foraging behavior of humpback whales. This technique is used to improve the parameters of xgboost and discover the most influential features. We evaluate the proposed model by using extensive datasets that include a diverse array of crops, soil compositions, climatic conditions, and geographic regions. The results suggest that the xgboost-WOA model outperforms traditional machine learning models in terms of both projected accuracy and efficiency. Furthermore, the suggested method showcases robust and reliable performance across different environmental circumstances, highlighting its potential for practical use in precision agriculture. This research emphasizes the effectiveness of combining AdaBoost with WOA for forecasting agricultural output. Furthermore, it contributes to the development of advanced predictive systems to support sustainable agricultural operations in adapting to climate variations and changing environmental conditions.