Your browser doesn't support javascript.
Model selection with decision support model for US natural gas consumption forecasting
Expert Systems with Applications ; 217, 2023.
Article in English | Scopus | ID: covidwho-2240865
ABSTRACT
Reliable prediction of natural gas consumption helps make the right decisions ensuring sustainable economic growth. This problem is addressed here by introducing a hybrid mathematical model defined as the Choquet integral-based model. Model selection is based on decision support model to consider the model performance more comprehensively. Different from the previous literature, we focus on the interaction between models when combine models. This paper adds grey accumulation generating operator to Holt-Winters model to capture more information in time series, and the grey wolf optimizer obtains the associated parameters. The proposed model can deal with seasonal (short-term) variability using season auto-regression moving average computation. Besides, it uses the long short term memory neural network to deal with long-term variability. The effectiveness of the developed model is validated on natural gas consumption due to the COVID-19 pandemic in the USA. For this, the model is customized using the publicly available datasets relevant to the USA energy sector. The model shows better robustness and outperforms other similar models since it consider the interaction between models. This means that it ensures reliable perdition, taking the highly uncertain factor (e.g., the COVID-19) into account. © 2023 Elsevier Ltd
Keywords

Full text: Available Collection: Databases of international organizations Database: Scopus Language: English Journal: Expert Systems with Applications Year: 2023 Document Type: Article

Similar

MEDLINE

...
LILACS

LIS


Full text: Available Collection: Databases of international organizations Database: Scopus Language: English Journal: Expert Systems with Applications Year: 2023 Document Type: Article