Bitte verwenden Sie diesen Link, um diese Publikation zu zitieren, oder auf sie als Internetquelle zu verweisen: https://hdl.handle.net/10419/251823 
Erscheinungsjahr: 
2022
Quellenangabe: 
[Journal:] Environmental Energy and Economic Research [ISSN:] 2676-4997 [Volume:] 6 [Issue:] 2 [Article No.:] S034 [Publisher:] Iranian Association for Energy Economics [Place:] Tehran [Year:] 2022
Verlag: 
Iranian Association for Energy Economics, Tehran
Zusammenfassung: 
Ensuring energy security is a major concern of policymakers and economic planners. This objective could be achieved by managing the energy supply and its demand. The latter has received less attention, especially in developing countries. Neglect of energy consumption and its accurate forecasting leads to potential outages and also unsustainable development. Nonlinear methods that are consistent with the nature of energy consumption have led to better results. Therefore, in the present study, both aspects of sustainable development in the determinants of energy demand and the nonlinear hybrid method have been used. We introduced a model based on sustainable development indicators to forecast energy consumption in Iran in which the relevant indicators are specified by the determination phase. To forecast energy consumption, we provided a new standard dataset for energy consumption in Iran (IREC) based on the data extracted from the World Bank and Ministry of Energy dataset in Iran. The highlight of this research is that it provided the most efficient features from the dataset using the genetic algorithm and five forecasting approaches based on machine learning methods. The algorithm was able to select 14 features as the most effective indicators in predicting energy consumption from all the 104 ones in the IREC with 500 repetitions. The empirical results indicated that the model can provide important indicators for energy consumption forecasting. The experiment result of the model using the GA-Based feature selection indicates that the hybrid model has had better results and GA-SVM and GA-MLP have the best result respectively.
Schlagwörter: 
Energy consumption prediction
Sustainable development
Predictive model
Machine learning
Data mining
Persistent Identifier der Erstveröffentlichung: 
Creative-Commons-Lizenz: 
cc-by Logo
Dokumentart: 
Article
Dokumentversion: 
Published Version
Erscheint in der Sammlung:

Datei(en):
Datei
Größe
304.84 kB





Publikationen in EconStor sind urheberrechtlich geschützt.