Feature Analysis and Predictive Modeling for Occupant-Centric Thermal Comfort in Naturally Ventilated University Cafeterias Using Advanced Machine Learning Models

DOI

10.2991/978-94-6239-668-5_46How to use a DOI?

Keywords

Thermal Comfort; Naturally Ventilated University Cafeterias; Feature Analysis; Machine Learning Models

Abstract

Thermal comfort is a critical aspect of outdoor environmental quality that significantly influences human health, well-being, and productivity, particularly in densely populated cities such as Dhaka, Bangladesh. Traditional predictive methods typically rely on simplified models and limited parameters, constraining their ability to capture the complex interactions between environmental factors and occupant behavior. Therefore, this study develops advanced machine learning models for feature analysis and predictive modeling to predict occupant-centric thermal comfort in naturally ventilated university cafeterias. A total of 500 samples were collected from two cafeterias during the summer season (June 2025 to August 2025). The dataset comprises 18 features covering environmental parameters, demographic data, human perception related, and architectural design characteristics. Random Forest (RF) and XGBoost were applied. Feature importance was evaluated using the Mean Decrease Impurity (MDI) approach, which estimates each feature’s contribution based on the average reduction in impurity achieved across all splits within the ensemble models. Additionally, Permutation-based feature importance was employed to evaluate each variable’s predictive contribution by measuring the increase in model error after randomly shuffling its values. Furthermore, SHapley Additive exPlanations (SHAP) analysis was used to interpret feature importance and increase model transparency. The findings consistently highlighted the importance of certain features in the analysis, such as Wind_Velocity, Humidity, Sweating_Effect etc. in predicting occupant comfort levels. This research provides critical insights into optimizing outdoor thermal comfort in naturally ventilated cafeterias, empowering evidence-based decisions in both architectural design and facility management. Future research should include broader datasets and multi-seasonal measurements to improve model generalization.

Copyright

© 2026 The Author(s)

Open Access

Open Access This chapter is licensed under the terms of the Creative Commons Attribution-NonCommercial 4.0 International License (http://creativecommons.org/licenses/by-nc/4.0/), which permits any noncommercial use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.

Download article (PDF)

TY  - CONF
AU  - Mim Mony
AU  - Nowshin Nowyer Oyshi
AU  - Lamia Jamal
AU  - Firuja Tasneem
AU  - Mahabuba Porosh Moni
AU  - Mohammad Nyme Uddin
PY  - 2026
DA  - 2026/05/14
TI  - Feature Analysis and Predictive Modeling for Occupant-Centric Thermal Comfort in Naturally Ventilated University Cafeterias Using Advanced Machine Learning Models
BT  - Proceedings of the International Conference on Current Problems in Engineering and Applied Sciences (ICCPEAS 2025)
PB  - Atlantis Press
SP  - 447
EP  - 456
SN  - 2352-5401
UR  - https://doi.org/10.2991/978-94-6239-668-5_46
DO  - 10.2991/978-94-6239-668-5_46
ID  - Mony2026
ER  -