Scholarlink

In the realm of educational infrastructure, ensuring optimal human comfort within built environments stands as a cornerstone for enhancing productivity and fostering well-being. This study delves into the sophisticated development of predictive models, leveraging Artificial Intelligence (AI) techniques, particularly machine learning (ML) algorithms, to classify thermal comfort levels both indoors and outdoors in an educational building located in Dhaka, Bangladesh. By prioritizing natural ventilation systems over mechanical counterparts, this research aims to deepen the understanding of the complex interplay between occupants, predominantly students, and their surrounding environment, with a focus on identifying influential features through SHAP (SHapley Additive exPlanations) analysis. Drawing upon a diverse dataset encompassing 12 relevant variables, including student comfort metrics, environmental parameters (such as temperature, humidity, indoor/outdoor CO2 levels), demographic details (age, gender, study level), urban landscape characteristics, and building attributes, the study has achieved remarkable predictive accuracy exceeding 80% using ML models. Notably, factors like occupant sitting surface, CO2 levels, gender, sitting orientation, and clothing choices have emerged as highly influential, overshadowing other variables in their impact on occupant comfort within the university setting. These findings underscore the pivotal roles of these features in shaping the comfort experiences of students. By shedding light on these key determinants, this study not only significantly advances the domain of thermal comfort analysis among university students but also offers invaluable insights crucial for the development of sustainable built environments within educational contexts. Through these revelations, the study contributes to a deeper understanding of how to optimize human comfort in educational settings, paving the way for more effective design and management strategies that prioritize the well-being and productivity of occupants.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Enhancing Human Thermal Comfort Prediction Models in Educational Buildings: An Artificial Intelligence (AI) Approach for Natural Ventilation Environments

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Mohammad Nyme Uddin,
  • Md. Tafhimul Hasan,
  • G. M. Rahad,
  • Hasibul Hasan Shawon,
  • Jobayer Hossain,
  • Md. Faisal Rahman Dhrubo,
  • Md. Hasan,
  • Md. Estiak Ahmed,
  • Humayun Kabir,
  • Md. Hasibul Islam

摘要

In the realm of educational infrastructure, ensuring optimal human comfort within built environments stands as a cornerstone for enhancing productivity and fostering well-being. This study delves into the sophisticated development of predictive models, leveraging Artificial Intelligence (AI) techniques, particularly machine learning (ML) algorithms, to classify thermal comfort levels both indoors and outdoors in an educational building located in Dhaka, Bangladesh. By prioritizing natural ventilation systems over mechanical counterparts, this research aims to deepen the understanding of the complex interplay between occupants, predominantly students, and their surrounding environment, with a focus on identifying influential features through SHAP (SHapley Additive exPlanations) analysis. Drawing upon a diverse dataset encompassing 12 relevant variables, including student comfort metrics, environmental parameters (such as temperature, humidity, indoor/outdoor CO2 levels), demographic details (age, gender, study level), urban landscape characteristics, and building attributes, the study has achieved remarkable predictive accuracy exceeding 80% using ML models. Notably, factors like occupant sitting surface, CO2 levels, gender, sitting orientation, and clothing choices have emerged as highly influential, overshadowing other variables in their impact on occupant comfort within the university setting. These findings underscore the pivotal roles of these features in shaping the comfort experiences of students. By shedding light on these key determinants, this study not only significantly advances the domain of thermal comfort analysis among university students but also offers invaluable insights crucial for the development of sustainable built environments within educational contexts. Through these revelations, the study contributes to a deeper understanding of how to optimize human comfort in educational settings, paving the way for more effective design and management strategies that prioritize the well-being and productivity of occupants.