Title: Regional steel price index predictions for North China through machine learning

Authors: Bingzi Jin; Xiaojie Xu

Addresses: Advanced Micro Devices (China) Co., Ltd., Shanghai, China ' North Carolina State University, Raleigh, NC 27695, USA

Abstract: Projections of commodity prices have long been heavily relied upon by investors and the government. This study examines the challenging task of estimating the daily regional steel price index in the north Chinese market for the period of 1 January 2010 to 15 April 2021. The projection of this significant commodity price indication has not received enough attention in the literature. After training our models with cross-validation and Bayesian optimisations, we apply Gaussian process regressions to verify our findings. The models that were built properly predicted the price indices between 8 January 2019 and 15 April 2021, with an out-of-sample relative root mean square error of 0.5871%. Investors and government officials can utilise the generated models for price analysis and decision-making. Forecasting results can help create comparable commodity price indices when reference data on the price trends suggested by these models are used.

Keywords: regional steel price index; time-series forecast; Gaussian process regression; GPR; Bayesian optimisation; cross validation; China.

DOI: 10.1504/IJMME.2024.140697

International Journal of Mining and Mineral Engineering, 2024 Vol.15 No.3, pp.314 - 350

Received: 10 Mar 2024
Accepted: 03 Jun 2024
Published online: 30 Aug 2024 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article