A Novel Graph based Deep Learning Model for Real Estate Finance

We aim to explore how urban green spaces, particularly parks, influence house prices in Guangzhou City by examining the association mechanism. With the advancement of urbanization, green Spaces are becoming increasingly important in urban development, especially due to their significant impact on property values. We collected data on green space and house prices in Tianhe, Guangzhou, including 12 parks and 1082 house price samples. We proposed a graph convolution network-based method for regression analysis and relationship extraction. The experiment shows that our model is the state-of-the-art model in home values regression task. We have shown that the larger the area of green space and the closer it is to green space, the higher the house price. By examining the complex relationship between urban green space and property values, this study provides valuable resources for urban planning and development.