Hierarchical and K-means clustering to assess thermal dissatisfaction and productivity in university classrooms

Classification of thermal environment control indicators according to the thermal sensitivity of office occupants

“…In addition to the area of thermal comfort, cluster analysis is widely used in medicine, information technology [ [53] , [54] , [55] ]. Despite all this research in different areas, studies involving thermal environment are few, and there are no studies applying cluster analysis [ 56 ]. In case, the clustering results can synthesize the characteristics of the occupant class based on appropriate attributes [ 57 ].…”

Comparative Analysis of Productivity Methods to Evaluate Thermal Comfort in Classrooms through Hierarchical Clustering

“…This helped to improve the organization and structuring of the stages, resulting in the overall procedure optimization, as highlighted by Zamboni et al [19]. After thoroughly analyzing the remaining studies, models that used parameters in line with the data applied in this study were chosen, specifically those collected during a recent IEQ Lab study developed by Bueno et al [13].…”

“…As mentioned above, the selected models were applied to the data from Bueno et al [13], which were collected between 23 March and 14 June 2022 during the fall in classrooms with natural ventilation, which held 42 students in 65 m 2 . A total of 519 valid responses were obtained through questionnaires and measurements taken in the morning, afternoon, and evening with a device similar to the BABUC-A microclimate station manufactured by Brüel and Kjaer, which monitored dry and wet bulb temperature, relative humidity, and globe temperature in real-time.…”

“…This study combines hierarchical cluster analysis and the k-means algorithm's application but aims to examine the relationship between thermal dissatisfaction and student productivity. The approach proposed in this paper differs from the one suggested in [13] since the previous study did not use specific productivity models to categorize them based on their similarities and differences to help fill this knowledge gap previously identified by Appel-Meulenbroek, Clippard, and Pfnür [14].…”

“…This approach naturally organizes clusters through branching, combining similar data sets [12]. A hierarchical cluster analysis was carried out based on data from the study by Bueno et al [13] in classrooms with natural ventilation at a Brazilian university that was previously investigated. This study combines hierarchical cluster analysis and the k-means algorithm's application but aims to examine the relationship between thermal dissatisfaction and student productivity.…”