 This research aimed to develop a ground subsidence risk prediction model based on machine learning. The model was trained using various datasets, including attribute information historical ground subsidence data from six different types of underground utilities. The model was then tested on a risk map of ground subsidence in the target area, and the results indicated that density was the most important factor affecting ground subsidence risk. This article was authored by Sun Ji-oh Lee, Jamo Kang, and Jin Young Kim.