Prediction of Temperature and Humidity at Telkom University Landmark Tower (TULT) Using Support Vector Machine (SVM) and Internet of Things (IoT)

Tall buildings tend to have a complex temperature management system. Telkom University Landmark Tower (TULT) is a building with varying floor heights that can significantly impact temperature and humidity levels. These variations directly affect the thermal comfort experienced by students, thus influencing the overall quality of teaching and learning activities. To address the challenge of differing conditions on each floor, this study aims to develop a temperature and humidity prediction system by combining the Support Vector Machine (SVM) method with the Internet of Things (IoT). The primary goal is to accurately forecast humidity and temperature at different elevations inside the building. In this research, the SVM method accompanied by IoT is employed to assess students’ thermal comfort during their activities at TULT. The system’s predictions will provide valuable insights into the adequacy of thermal comfort, facilitating any necessary improvements. The results obtained from measurements and evaluations demonstrate the effectiveness of the SVM method in humidity and temperature prediction. Notably, the SVM method yields relatively low error predicted values. For instance, during the training phase, the SVM method achieved a Mean Absolute Error (MAE) of 0.269 for humidity prediction and 0.04 for temperature prediction. In the testing phase, the SVM method exhibited a slightly higher MAE of 0.321 for humidity and 0.107 for temperature. The small MAE values suggest that the SVM method performs well in providing accurate predictions.