Performance of Machine Learning Algorithms considering Spatial Effects Assessment for Indoor Personal Thermal Comfort in Air-Conditioned Workplace

Personal comfort models were developed to circumvent most of the constraints imposed by Predicted Mean Vote ( PMV ) and present adaptive models, which consider average response a large population. Although there has been lot research into new input features for personal spatial data building, such as windows, doors, furniture, walls, fans, heating, ventilation, air conditioning HVAC systems, (the location its occupants with those elements), have not thoroughly examined. This paper investigates impact parameter in predicting indoor thermal using various machine learning approaches air-conditioning offices under hot humid climates. The Decision Tree, Random Forest, Support Vector Machine, K-Nearest Neighbour, Neural Network trained field study dataset that was done nineteen office spaces yielding 628 samples from 42 occupants. is divided randomly training testing datasets, ratio 80% 20%. examines how well predicts compared without data; where parameters shown significant influence on model prediction accuracies, Absolute Error MAE ), Root Squared RMSE ). result shows decreased 10.6% Forest (RF) getting reduction 23.8%. Meanwhile, reduced 11.8% RF giving cutback 30.6%. Consequently, effect analysis also determines area room cold or heat clusters affects contributes design sustainable buildings.