Advertisement

IRC-SET 2018 pp 363-373 | Cite as

Predicting Individual Thermal Comfort

  • Lim Xin YiEmail author
  • Lee Jia Jia
  • Daren Ler
Conference paper

Abstract

Thermal comfort is a very important factor in many people’s lives; when people do not feel comfortable, their focus, productivity and performance are affected. In our research, we investigate how to help students achieve optimal comfort by predicting their comfort levels. We collected data on 5 students, and then utilised the k-nearest neighbour machine learning algorithm, in conjunction with tenfold cross-validation, to generate models of student comfort in the classroom. The features utilised are air temperature, air velocity, air relative humidity, body temperature and heart rate. In our experiments, we seek to learn if acceptable individual models may be derived, and more importantly, if combined models can help increase predictive accuracy. Our work suggests that combined models are applicable only when used to augment datasets that are applied to a subject with a more complex thermal comfort model.

Keywords

k-nearest neighbor algorithm Thermal comfort Machine learning Accuracy rate Variables Features Relative humidity Heart rate Wind speed Body temperature Time of the day Air temperature Self-assessed comfort level Comfort level k value Individual subjects Merged dataset Datasets Classifier Tenfold cross validation Wrapper based technique Instances 

References

  1. 1.
    Cover, T., & Hart, P. (1967). Nearest neighbor pattern classification. IEEE Transactions on Information Theory, 13(1), 21–27.CrossRefGoogle Scholar
  2. 2.
    Fanger, P. O. (1970). Thermal comfort. Analysis and applications in environmental engineering.Google Scholar
  3. 3.
    Hall, M., Frank, E., Holmes, G., Pfahringer, B., Reutemann, P., & Witten, I. H. (2009). The WEKA data mining software: An update. ACM SIGKDD Explorations Newsletter, 11(1), 10–18.CrossRefGoogle Scholar
  4. 4.
    Stone, M. (1974). Cross-validatory choice and assessment of statistical predictions. Journal of the Royal Statistical Society. Series B (Methodological), 111–147.Google Scholar
  5. 5.
    Laftchiev, E., & Nikovoski, D. N. (2016, December). An IoT system to estimate personal thermal comfort. Retrieved from https://www.merl.com/publications/docs/TR2016-161.PDF.
  6. 6.
    Choi, J.-H., Loftness, V., & Lee, D.-W. (2011, October). Investigation of the possibility of the use of heart rate as a human factor for thermal sensation models. Retrieved from https://viterbipk12.usc.edu/wp-content/uploads/2017/06/2012-Building-and-Environment-Choi2c-Loftness2c-Lee-Investigation-of-the-possibility-of-the-use-of-heart-rate-as-a-human-factor-for.PDF.
  7. 7.
    Rabadi, N. J. (2011, August). Developing a software to predict thermal comfort of humans at work. Retrieved from http://jjmie.hu.edu.jo/files/v5n4/JJMIE-116-11.PDF.

Copyright information

© Springer Nature Singapore Pte Ltd. 2019

Authors and Affiliations

  1. 1.National Junior CollegeSingaporeSingapore

Personalised recommendations