Predicting total household energy expenditures using ensemble learning methods

Esma Kesriklioğlu, Erkan Oktay and Abdulkerim Karaaslan

Energy, 2023, vol. 276, issue C

Abstract: Total household energy expenditures are a complex topic because so many behavioral, technological, environmental, and policy variables can affect expenditures. This study aimed to develop a high-performance ensemble learning (EL) model to classify total household energy expenditures. For this purpose, household consumption data from 11,521 households were examined using the Household Budget Survey 2019 data set that the Turkish Statistical Institute (TURKSTAT) published. In addition to the variables directly related to household energy expenditures, new variables were created within the framework of the literature and under the guidance of expert opinion. The prepared data were passed through data preprocessing, modeling, prediction, and performance evaluation stages using the open source RapidMiner software program. Classification performances of machine learning and EL methods were compared. Aside from k-nearest neighbor, decision tree, naive Bayes, random forest, gradient boosted trees, and DFNN classifiers, the study used bagging, boosting, voting, and stacking EL methods. The stacking EL method in the ALL model and bagging EL method in the deep feed forward neural network (DFNN) classifiers achieved the highest performance among EL methods. The accuracy value of the stacking and bagging methods was 0.984. The results indicate that EL methods can enhance individual machine learning methods significantly.

Keywords: Ensemble learning; Household energy expenditure; Classification; Machine learning; CRISP-DM (search for similar items in EconPapers)
Date: 2023
References: View references in EconPapers View complete reference list from CitEc
Citations: View citations in EconPapers (2)

Downloads: (external link)
http://www.sciencedirect.com/science/article/pii/S0360544223009751
Full text for ScienceDirect subscribers only

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:eee:energy:v:276:y:2023:i:c:s0360544223009751

DOI: 10.1016/j.energy.2023.127581

Access Statistics for this article

Energy is currently edited by Henrik Lund and Mark J. Kaiser

More articles in Energy from Elsevier
Bibliographic data for series maintained by Catherine Liu ().