Forecasting Natural Gas Spot Prices with Machine Learning
The ability to accurately forecast the spot price of natural gas benefits stakeholders and is a valuable tool for all market participants in the competitive gas market. In this paper, we attempt to forecast the natural gas spot price 1, 3, 5, and 10 days ahead using machine learning methods: support...
Main Authors: | , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
MDPI AG
2021-09-01
|
Series: | Energies |
Subjects: | |
Online Access: | https://www.mdpi.com/1996-1073/14/18/5782 |
id |
doaj-b0ad7718e89c470e93314ca81d19a17b |
---|---|
record_format |
Article |
spelling |
doaj-b0ad7718e89c470e93314ca81d19a17b2021-09-26T00:05:08ZengMDPI AGEnergies1996-10732021-09-01145782578210.3390/en14185782Forecasting Natural Gas Spot Prices with Machine LearningDimitrios Mouchtaris0Emmanouil Sofianos1Periklis Gogas2Theophilos Papadimitriou3Faculty of Sciences, Aristotle University of Thessaloniki, 54124 Thessaloniki, GreeceDepartment of Economics, Democritus University of Thrace, 69100 Komotini, GreeceDepartment of Economics, Democritus University of Thrace, 69100 Komotini, GreeceDepartment of Economics, Democritus University of Thrace, 69100 Komotini, GreeceThe ability to accurately forecast the spot price of natural gas benefits stakeholders and is a valuable tool for all market participants in the competitive gas market. In this paper, we attempt to forecast the natural gas spot price 1, 3, 5, and 10 days ahead using machine learning methods: support vector machines (SVM), regression trees, linear regression, Gaussian process regression (GPR), and ensemble of trees. These models are trained with a set of 21 explanatory variables in a 5-fold cross-validation scheme with 90% of the dataset used for training and the remaining 10% used for testing the out-of-sample generalization ability. The results show that these machine learning methods all have different forecasting accuracy for every time frame when it comes to forecasting natural gas spot prices. However, the bagged trees (belonging to the ensemble of trees method) and the linear SVM models have superior forecasting performance compared to the rest of the models.https://www.mdpi.com/1996-1073/14/18/5782natural gasspot pricemachine learningforecasting |
collection |
DOAJ |
language |
English |
format |
Article |
sources |
DOAJ |
author |
Dimitrios Mouchtaris Emmanouil Sofianos Periklis Gogas Theophilos Papadimitriou |
spellingShingle |
Dimitrios Mouchtaris Emmanouil Sofianos Periklis Gogas Theophilos Papadimitriou Forecasting Natural Gas Spot Prices with Machine Learning Energies natural gas spot price machine learning forecasting |
author_facet |
Dimitrios Mouchtaris Emmanouil Sofianos Periklis Gogas Theophilos Papadimitriou |
author_sort |
Dimitrios Mouchtaris |
title |
Forecasting Natural Gas Spot Prices with Machine Learning |
title_short |
Forecasting Natural Gas Spot Prices with Machine Learning |
title_full |
Forecasting Natural Gas Spot Prices with Machine Learning |
title_fullStr |
Forecasting Natural Gas Spot Prices with Machine Learning |
title_full_unstemmed |
Forecasting Natural Gas Spot Prices with Machine Learning |
title_sort |
forecasting natural gas spot prices with machine learning |
publisher |
MDPI AG |
series |
Energies |
issn |
1996-1073 |
publishDate |
2021-09-01 |
description |
The ability to accurately forecast the spot price of natural gas benefits stakeholders and is a valuable tool for all market participants in the competitive gas market. In this paper, we attempt to forecast the natural gas spot price 1, 3, 5, and 10 days ahead using machine learning methods: support vector machines (SVM), regression trees, linear regression, Gaussian process regression (GPR), and ensemble of trees. These models are trained with a set of 21 explanatory variables in a 5-fold cross-validation scheme with 90% of the dataset used for training and the remaining 10% used for testing the out-of-sample generalization ability. The results show that these machine learning methods all have different forecasting accuracy for every time frame when it comes to forecasting natural gas spot prices. However, the bagged trees (belonging to the ensemble of trees method) and the linear SVM models have superior forecasting performance compared to the rest of the models. |
topic |
natural gas spot price machine learning forecasting |
url |
https://www.mdpi.com/1996-1073/14/18/5782 |
work_keys_str_mv |
AT dimitriosmouchtaris forecastingnaturalgasspotpriceswithmachinelearning AT emmanouilsofianos forecastingnaturalgasspotpriceswithmachinelearning AT periklisgogas forecastingnaturalgasspotpriceswithmachinelearning AT theophilospapadimitriou forecastingnaturalgasspotpriceswithmachinelearning |
_version_ |
1717367144933490688 |