Various multistage ensembles for prediction of heating energy consumption
Abstract
Feedforward neural network models are created for prediction of daily heating energy consumption of a NTNU university campus Gloshaugen using actual measured data for training and testing. Improvement of prediction accuracy is proposed by using neural network ensemble. Previously trained feed-forward neural networks are first separated into clusters, using k-means algorithm, and then the best network of each cluster is chosen as member of an ensemble. Two conventional averaging methods for obtaining ensemble output are applied; simple and weighted. In order to achieve better prediction results, multistage ensemble is investigated. As second level, adaptive neuro-fuzzy inference system with various clustering and membership functions are used to aggregate the selected ensemble members. Feedforward neural network in second stage is also analyzed. It is shown that using ensemble of neural networks can predict heating energy consumption with better accuracy than the best trained single neu...ral network, while the best results are achieved with multistage ensemble.
Keywords:
multistage neural network ensemble / heating consumption prediction / adaptive neuro-fuzzy inferenceSource:
Modeling Identification and Control, 2015, 36, 2, 119-132Publisher:
- Mic, Trondheim
Funding / projects:
- Norwegian Programme in Higher Education, Research and Development in the Western Balkans, Programme 3: Energy Sector (HERD Energy)
DOI: 10.4173/mic.2015.2.4
ISSN: 0332-7353
WoS: 000359471300002
Scopus: 2-s2.0-84937030621
Collections
Institution/Community
Mašinski fakultetTY - JOUR AU - Jovanović, Radiša AU - Sretenović, Aleksandra PY - 2015 UR - https://machinery.mas.bg.ac.rs/handle/123456789/2092 AB - Feedforward neural network models are created for prediction of daily heating energy consumption of a NTNU university campus Gloshaugen using actual measured data for training and testing. Improvement of prediction accuracy is proposed by using neural network ensemble. Previously trained feed-forward neural networks are first separated into clusters, using k-means algorithm, and then the best network of each cluster is chosen as member of an ensemble. Two conventional averaging methods for obtaining ensemble output are applied; simple and weighted. In order to achieve better prediction results, multistage ensemble is investigated. As second level, adaptive neuro-fuzzy inference system with various clustering and membership functions are used to aggregate the selected ensemble members. Feedforward neural network in second stage is also analyzed. It is shown that using ensemble of neural networks can predict heating energy consumption with better accuracy than the best trained single neural network, while the best results are achieved with multistage ensemble. PB - Mic, Trondheim T2 - Modeling Identification and Control T1 - Various multistage ensembles for prediction of heating energy consumption EP - 132 IS - 2 SP - 119 VL - 36 DO - 10.4173/mic.2015.2.4 ER -
@article{ author = "Jovanović, Radiša and Sretenović, Aleksandra", year = "2015", abstract = "Feedforward neural network models are created for prediction of daily heating energy consumption of a NTNU university campus Gloshaugen using actual measured data for training and testing. Improvement of prediction accuracy is proposed by using neural network ensemble. Previously trained feed-forward neural networks are first separated into clusters, using k-means algorithm, and then the best network of each cluster is chosen as member of an ensemble. Two conventional averaging methods for obtaining ensemble output are applied; simple and weighted. In order to achieve better prediction results, multistage ensemble is investigated. As second level, adaptive neuro-fuzzy inference system with various clustering and membership functions are used to aggregate the selected ensemble members. Feedforward neural network in second stage is also analyzed. It is shown that using ensemble of neural networks can predict heating energy consumption with better accuracy than the best trained single neural network, while the best results are achieved with multistage ensemble.", publisher = "Mic, Trondheim", journal = "Modeling Identification and Control", title = "Various multistage ensembles for prediction of heating energy consumption", pages = "132-119", number = "2", volume = "36", doi = "10.4173/mic.2015.2.4" }
Jovanović, R.,& Sretenović, A.. (2015). Various multistage ensembles for prediction of heating energy consumption. in Modeling Identification and Control Mic, Trondheim., 36(2), 119-132. https://doi.org/10.4173/mic.2015.2.4
Jovanović R, Sretenović A. Various multistage ensembles for prediction of heating energy consumption. in Modeling Identification and Control. 2015;36(2):119-132. doi:10.4173/mic.2015.2.4 .
Jovanović, Radiša, Sretenović, Aleksandra, "Various multistage ensembles for prediction of heating energy consumption" in Modeling Identification and Control, 36, no. 2 (2015):119-132, https://doi.org/10.4173/mic.2015.2.4 . .