A Bayesian Model for Optimizing Thermal Comfort and Indoor Air Quality

Akdoğan M., Alaçam S., Töreyin B. U.

Accelerated Landscapes: XXVII SIGraDi (Sociedad Iberoamericana de Gráfica Digital) Conference 2023, Maldonado, Uruguay, 29 November - 01 December 2023, pp.1-12

  • Publication Type: Conference Paper / Full Text
  • City: Maldonado
  • Country: Uruguay
  • Page Numbers: pp.1-12
  • Istanbul Technical University Affiliated: Yes


This study focuses on the usage of a basic probabilistic approach on determining the best course of action in a specific environment within the domain of architecture. More specifically, Bayesian decision theory is applied on a basic level for the problem of maintaining thermal comfort and air quality. An already existing comprehensive dataset is used and narrowed down for the purpose of the study. Environment measurements (indoor and outdoor temperature, indoor CO2 level and air humidity) are taken as input variables and user preferences (open or closed window) are taken as outputs in order to address the problem as a binary classification problem. The study can be regarded as a basic example on the usage of probabilistic approaches in the discipline of architecture.