A data mining-based daylighting design decision support model for achieving visual comfort conditions in the multi-functional residential space

The rise of remote working has transformed residential spaces into multi-functional environments, accommodating working, living, and resting activities. In parallel, unlike the traditional residential spaces, the new-era multi-functional residential spaces (MFRs) require the daylighting systems that must address diverse and simultaneous visual comfort requirements arising from various activities. To meet these complex needs, this study aims to develop a daylighting design decision model for MFRs to optimize residents’ comfort, health, and performance across various visual tasks. In this regard, a novel Curriculum Decision Tree (curriculum DT) model based on data mining is purposed to support daylighting design decision depending on the visual comfort level in multi-functional residential spaces by considering different design stages. The model has been tested using a parametric approach and curriculum DT model for a generic residential space in Istanbul, representing a Mediterranean climate. The model identifies orientation-based daylighting design paths tailored to daylight availability, glare, view-out, and sunlight exposure. The proposed model also reveals that, in addition to orientation, window width and shading device depth are the key parameters affecting visual comfort level in the multi-functional residential spaces. From a data mining perspective, the proposed curriculum decision tree reduces complexity, optimizes feature placement, achieves a predictive accuracy of 0.81, recall of 0.81, and precision of 0.82, serving as a reliable guide for building professionals in addressing real design problems. The outputs of the study can provide valuable insights into daylighting design approaches for MFRs, offering strategies to enhance visual comfort and support holistic well-being and productivity.