Change search
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • apa
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf
An Advanced Explainable Belief Rule-Based Framework to Predict the Energy Consumption of Buildings
Luleå University of Technology, Department of Computer Science, Electrical and Space Engineering, Computer Science.ORCID iD: 0000-0001-5283-6641
Department of Computer Science & Engineering, University of Chittagong, Chattogram 4331, Bangladesh.ORCID iD: 0000-0002-7473-8185
Luleå University of Technology, Department of Computer Science, Electrical and Space Engineering, Computer Science.ORCID iD: 0000-0003-0244-3561
2024 (English)In: Energies, E-ISSN 1996-1073, Vol. 17, no 8, article id 1797Article in journal (Refereed) Published
Abstract [en]

The prediction of building energy consumption is beneficial to utility companies, users, and facility managers to reduce energy waste. However, due to various drawbacks of prediction algorithms, such as, non-transparent output, ad hoc explanation by post hoc tools, low accuracy, and the inability to deal with data uncertainties, such prediction has limited applicability in this domain. As a result, domain knowledge-based explainability with high accuracy is critical for making energy predictions trustworthy. Motivated by this, we propose an advanced explainable Belief Rule-Based Expert System (eBRBES) with domain knowledge-based explanations for the accurate prediction of energy consumption. We optimize BRBES’s parameters and structure to improve prediction accuracy while dealing with data uncertainties using its inference engine. To predict energy consumption, we take into account floor area, daylight, indoor occupancy, and building heating method. We also describe how a counterfactual output on energy consumption could have been achieved. Furthermore, we propose a novel Belief Rule-Based adaptive Balance Determination (BRBaBD) algorithm for determining the optimal balance between explainability and accuracy. To validate the proposed eBRBES framework, a case study based on Skellefteå, Sweden, is used. BRBaBD results show that our proposed eBRBES framework outperforms state-of-the-art machine learning algorithms in terms of optimal balance between explainability and accuracy by 85.08%.

Place, publisher, year, edition, pages
MDPI, 2024. Vol. 17, no 8, article id 1797
Keywords [en]
accuracy, building energy, explainability, trust, uncertainties
National Category
Computer Sciences
Research subject
Cyber Security; Pervasive Mobile Computing
Identifiers
URN: urn:nbn:se:ltu:diva-105059DOI: 10.3390/en17081797OAI: oai:DiVA.org:ltu-105059DiVA, id: diva2:1850830
Funder
Vinnova, 2022-01188
Note

Validerad;2024;Nivå 2;2024-04-11 (signyg);

Full text license: CC BY

Available from: 2024-04-11 Created: 2024-04-11 Last updated: 2024-04-24Bibliographically approved
In thesis
1. A Novel Explainable Belief Rule-Based Prediction Framework under Uncertainty
Open this publication in new window or tab >>A Novel Explainable Belief Rule-Based Prediction Framework under Uncertainty
2024 (English)Licentiate thesis, comprehensive summary (Other academic)
Abstract [en]

Traditional Machine Learning (ML) and Deep Learning (DL) models provide very accurate predictions because of their intricate mathematical operations. However, these models do not explain the reasons in support of the predictive outputs. Therefore, there is no trust between humans and AI when it produces such obtuse results. The widespread adoption of AI models heavily relies on the trust that humans place in the decisions made by AI. This trust holds great importance in safety-critical sectors like healthcare, autonomous vehicles, and the energy domain. Several post-hoc tools can help elucidate the results of AI models. The use of training datasets rather than domain knowledge renders such an explanation as a proxy. A misleading explanation will be produced by a biased training dataset. In contrast, an end user is more likely to believe an explanation that is grounded in domain knowledge. Motivated by this, we propose a novel framework, consisting of Belief Rule Based Expert System (BRBES), to predict output and explain it with reference to domain knowledge. BRBES effectively utilizes its rule base to represent domain knowledge and adeptly handles uncertainty resulting from a lack of information. We also fine-tune the parameters and structure of BRBES to enhance the accuracy of the prediction. Therefore, the output of our proposed framework is not only accurate, but also easily explainable.

This licentiate thesis delves into the challenges and opportunities surrounding eXplainable ArtificialIntelligence (XAI) in order to provide a comprehensive understanding of AI output. It presents a new XAI framework that can effectively predict an output and provide explanations based on domain knowledge. In addition, it shows how effective BRBES integration can be by integrating it with a deep learning model to forecast air quality phenomenon using ground and satellite data.

This thesis presents four significant contributions. First, we conduct a comprehensive examination ofthe existing literature on XAI, delving into the numerous challenges and opportunities it offers. Extensive research has been conducted to explore the definition, classification, and practical use of XAI. In addition, this complex study highlights the significance of the user interface in effectively communicating explanations in a way that is comprehensible to humans. It also brings up the tradeoff between explainability and accuracy in XAI. Secondly, we present an innovative explainable BRBES (eBRBES) model that offers accurate predictions of building energy consumption phenomenon while providing insightful explanations based on domain knowledge. As part of eBRBES, we also present a novel Belief Rule Based adaptive Balance Determination (BRBaBD) algorithm to assess the optimal balance between explainability and accuracy. Thirdly, we propose a mathematical model to integrate BRBES with the Convolutional Neural Network (CNN). We leverage the domain knowledge of BRBES, and hidden data patterns discovered by CNN with this integrated approach. We predict air quality withthis integrated model using outdoor ground images and sensor data. Fourth, we integrate two-layer BRBES with CNN to monitor air quality using satellite images, and relevant environmental parameters, such as cloud, relative humidity, temperature, and wind speed. The two-layer BRBES showcases the strength of BRBES in conducting reasoning across multiple layers.

Based on the research findings of this thesis applied on two different phenomena, it can be argued that utilizing a belief rule-based framework can enhance predictability with greater clarity and precision. 

Place, publisher, year, edition, pages
Luleå: Luleå University of Technology, 2024
Series
Licentiate thesis / Luleå University of Technology, ISSN 1402-1757
National Category
Computer Sciences
Research subject
Pervasive Mobile Computing
Identifiers
urn:nbn:se:ltu:diva-105200 (URN)978-91-8048-554-8 (ISBN)978-91-8048-555-5 (ISBN)
Presentation
2024-06-20, A193, Luleå University of Technology, Skellefteå, 09:00 (English)
Opponent
Supervisors
Funder
Vinnova, 2022-01188
Note

Funder: Rönnbäret Foundation, Sweden

Available from: 2024-04-24 Created: 2024-04-24 Last updated: 2024-05-30Bibliographically approved

Open Access in DiVA

fulltext(2721 kB)31 downloads
File information
File name FULLTEXT01.pdfFile size 2721 kBChecksum SHA-512
f8efa86f503fe36086884458fcbcb0eccfbd1d7a09831aa87dfc273f92f53f6e4b611385d58fecaa1949ac6ef79306f8a4e0b75a40c80b8b3bb47cec15badbfe
Type fulltextMimetype application/pdf

Other links

Publisher's full text

Authority records

Kabir, SamiAndersson, Karl

Search in DiVA

By author/editor
Kabir, SamiHossain, Mohammad ShahadatAndersson, Karl
By organisation
Computer Science
In the same journal
Energies
Computer Sciences

Search outside of DiVA

GoogleGoogle Scholar
Total: 31 downloads
The number of downloads is the sum of all downloads of full texts. It may include eg previous versions that are now no longer available

doi
urn-nbn

Altmetric score

doi
urn-nbn
Total: 412 hits
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • apa
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf