대한건축학회논문집 (Journal of the Architectural Institute of Korea)

  • 제40권3호
  • /
  • Pages.81-88
  • /
  • 2024
  • /
  • 2733-6239(pISSN)
  • /
  • 2733-6247(eISSN)
대한건축학회 (Architectural Institute of Korea)
권호 표지
DOI QR코드

DOI QR Code

탐색적 데이터 분석 방법을 활용한 머신러닝 기반 재실인원 추정 모델

Discovering Anomalous Power Usage Patterns in Rental Housing Through Small-Scale Data

  • 투고 : 2023.12.13
  • 심사 : 2024.01.27
  • 발행 : 2024.03.30
⟨ 이전 논문 다음 논문 ⟩

초록

Employing machine learning, this study detected occupancy anomalies in Seoul's public rental housing by analyzing energy usage data spanning from 2016 to 2021. Through the examination of electricity consumption patterns, the model successfully pinpointed instances of underreported or illegal occupancy, identifying approximately 8% of households as anomalies. This approach highlights the promise of data-driven methodologies in public housing management, promoting adherence to regulations and equitable resource allocation. Visualization of results using GIS further facilitates their practical utilization by housing authorities.

키워드

참고문헌

  1. Aliero, M. S., Qureshi, K. N., Pasha, M. F., & Jeon, G. (2021). Smart Home Energy Management Systems in Internet of Things networks for green cities demands and services. Environmental Technology & Innovation, 22, 101443.
  2. Alzoubi, A. (2022). Machine learning for intelligent energy consumption in smart homes. International Journal of Computations, Information and Manufacturing (IJCIM), 2(1).
  3. Bae, W. B., Ko, J. H., Yeo, J. E., Mun, S. H., & Huh, J. H. (2015). An Occupants Location-based Plugin Device Control. Journal of the Architectural Institute of Korea Conference Proceedings, 397-398.
  4. Chen, C. F., de Rubens, G. Z., Xu, X., & Li, J. (2020). Coronavirus comes home? Energy use, home energy management, and the social-psychological factors of COVID-19. Energy Research & Social Science, 68, 101688.
  5. Choi, B. S., & Park, J. A. (2019). A Study on the Actual Condition of Residential Environment in an Apartment with a High Ratio of the Aged in Seoul from the Viewpoint of the NORC - Focused on Apartments in Nowon-gu and Gangnam-gu Areas. Journal of the Korean Housing Association, 30(4), 21-30.
  6. Cho, Y. J., & Kim, J. H. (2021). Data analysis for Load forecasting of buildings using machine learning methods. Transactions of the Korean Institute of Electrical Engineers, 750-751.
  7. Dixon, W. J., & Massey, F. J. (1951). Introduction to statistical analysis. McGraw-Hill.
  8. Han, H, T., & Han, C. H. (2012). Occupancy Estimation based on CO2 Concentrations considering Disturbances of Vicinity Zones. Proceedings of the Society of Air-Conditioning and Refrigerating Engineers of Korea Conference, 139-142.
  9. Han, J. H., & Nam, J. (2023). The Characteristics of Public Housing in Seoul and the Economic Improvement of Households. Korean Urban Planning and Land Institute, 58(2), 104-115.
  10. Huh, M. H., & Jung, J. H. (1990). Software Review of Statistical Package Programs on EDA Aspects. Communications for Statistical Applications and Methods, 17 - 25.
  11. Jung, B. J. (2019). Consumer Perception Survey on Smart Meter and Meter Management in Apartments. Korean Institute of Electrical Engineers Conference, 504-505.
  12. Kang, H. M., & Kang S. J. (2021). Indoor movement pattern recognition using embedded machine learning system and 2-dimensional infrared array sensor. Proceedings of the Korea Information Science Society Conference, 180-182.
  13. Kim, S. H., & Seo, D. H. (2017). A Study on Occupancy Estimation Method of a Private Room Using IoT Sensor Data Based Decision Tree Algorithm. Journal of the Korean Solar Energy Society, 37(2), 23-33. https://doi.org/10.7836/KSES.2017.37.2.023
  14. Maheswari, C., Priyanka, E. B., Thangavel, S., Vignesh, S. R., & Poongodi, C. (2020). Multiple regression analysis for the prediction of extraction efficiency in mining industry with industrial IoT. Production Engineering, 14, 457-471.
  15. Park, C. H., & Kim, T. (2020). Energy theft detection in advanced metering infrastructure based on anomaly pattern detection. Energies, 13(15), 3832.