ANALYSING THE THERMAL COMFORT OF A COFFEE SHOP IN SINGAPORE

Abstract

No known literature which existed examine thermal comfort in coffee shops in Singapore before. It is pertinent that such study be conducted owning to the backdrop of climate change and countries in the tropics are expected to experience the highest increase in temperature which would affect the thermal comfort of occupants in different outdoor environment. This study aims to provide a comprehensive insight and overview into the factors at play which affect thermal comfort of respondents based on data collected through objective and subjective measurement in a coffee shop. The parameters are analysed and ranked using linear, multivariate linear regression and the supervised machine learning method of Random forest to understand the significance of each variable in contributing to thermal comfort in the coffee shop. Results showed that a linear regression is not very effective in explaining thermal comfort of respondents based on the objective parameters of temperature and relative humidity. However, in determining variable importance, mean wind speed came out on top in predicting thermal comfort in coffee shop while humidity is the most important factor in explaining thermal acceptability. Future studies could divide the coffee shop into sections to analyse individually which factors contribute most to thermal comfort in each section to obtain an average score. Other types of regression analysis and machine learning methods could be considered and compared to see which type of regression and algorithm model is best suitable to predict thermal comfort of respondents in a coffee shop.