Forecasting on metal resource spot settlement price: New evidence from the machine learning model
Resources Policy
; 81:103360.0, 2023.
Article
in English
| ScienceDirect | ID: covidwho-2245156
ABSTRACT
Accurate prediction of the price of metal mineral resources is of great practical significance for guiding the production of non-renewable resource enterprises and maintaining the stability of related economic price indexes. Based on the daily frequency spot settlement price data of lead, aluminium, aluminium alloy, tin, copper(A) and other metal minerals from Sep. 21, 2005 to Dec. 1, 2021, machine learning method is used to analyze the accuracy of hybrid models for predicting spot settlement prices of metal minerals. Firstly, we compared and analyzed the in-sample prediction accuracy of different models in spot settlement price of metal minerals, and found that the prediction accuracy of LSTM-GRU and LSTM-CNN models is significantly better than other models. Secondly, we verified the out-of-sample prediction results of spot settlement prices of metal minerals, which further demonstrates the robustness of the prediction accuracy of the LSTM hybrid model. Finally, we considered the impact of COVID-19 and explored the prediction accuracy of different hybrid models on spot settlement prices of metal minerals. We found that LSTM-GRU and other models also perform well with strong robustness. Therefore, we believed that the LSTM hybrid model, especially the LSTM-GRU model, is suitable for analyzing the prediction of spot settlement price of metal minerals.
Full text:
Available
Collection:
Databases of international organizations
Database:
ScienceDirect
Language:
English
Journal:
Resources Policy
Year:
2023
Document Type:
Article
Similar
MEDLINE
...
LILACS
LIS