Urban heat islands (UHI) pose significant environmental and health challenges in rapidly urbanizing cities like Hanoi. This study presents an integrated approach utilizing machinelearning model and cloud computing to predict and delineate UHIs in Hanoi. Leveraging high-resolution satellite imagery, meteorological data, population dentistry data and land cover factors, we develop a robust predictive model that accurately identifies UHI- prone areas. The data preprocessing involves normalizing diverse data sources, handling missing values, and ensuring spatial-temporal alignment. Feature selection is performed to identify the most influential factors contributing to UHI, including land surface temperature, vegetation index, population density, and urban morphology. Results indicate that the machine learning models, particularly the ensemble methods, exhibit high predictive accuracy, with the RandomForest model achieving an R2 of 0,66. In conclusion, applying machine learning models and cloud computing presents a powerful framework for predicting and managing urban heat islands. The study's innovative approach enhances understanding of UHI phenomena in Hanoi and provides a scalable solution adaptable to other urban settings. Future research should focus on refining model accuracy by incorporating additional data sources and exploring the socioeconomic impacts of UHI mitigation strategies.