Reliable novel hybrid extreme gradient boosting for forecasting copper prices using meta-heuristic algorithms: A thirty-year analysis.

Forecasting copper price accurately is essential in various aspects of economics. However, due to the complex and unpredictable fluctuations in copper prices make this task quite challenging. Then, the goal of the current study is to develop a novel hybrid intelligence method, namely SSO-XGB and HHO-XGB models, to accurately forecast copper prices in the future by utilizing extreme gradient boosting (XGB) and three metaheuristic algorithms – sparrow search optimization (SSO) and harris hawk optimization (HHO). To do this, time series datasets of copper prices over a 30-year period (1993–2023) were collected, taking into account influencing factors including aluminum, natural gas, gold, crude oil, silver, iron ore, and nickel prices, as well as exchange rates of the most significant countries. The XGB model's hyperparameters were fine-tuned by the SSO and HHO algorithms to achieve accurate copper price forecasts. In addition, comparative models were developed using the traditional XGB (optimized by grid search) model and gene expression programming (GEP) to obtain convincing experimental results for the proposed SSO-XGB and HHO-XGB models. Among these, the SSO-XGB model demonstrated the highest level of accuracy, achieving an RMSE, AARE, R2, and MAE of 106, 4.59, 0.992, and 58 respectively on the training data. Additionally, the sensitivity analysis (measured through mutual information (MI)) revealed that changes in gold and crude oil prices significantly impacted the copper price. In general, all four intelligence methods evaluated in this study demonstrated an acceptable level of performance for predicting copper prices. • Novel hybrid intelligence models: Introduces SSO-XGB and HHO-XGB models for accurate copper price forecasting. • Performance assessment: Comparative analysis identifies SSO-XGB as the most accurate model for copper price prediction. • Practical implications: Findings have practical implications for businesses, providing reliable copper price forecasts. • Evaluation of multiple models: Comparative analysis assesses the effectiveness of various models in predicting copper prices. [ABSTRACT FROM AUTHOR]

Copyright of Resources Policy is the property of Pergamon Press - An Imprint of Elsevier Science and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)