No abstract

This paper takes the Beijing's main urban as the research object, utilizing machine learning method to quantitatively simulate the location and layout of tourism and leisure facilities. POI data on the grid scale of 1km and the population data of the "the 6th Census" were used, combined with the existing distribution characteristics of tourism and leisure facilities in the Beijing's main urban. Finally, the model is used to simulate and predict each grid in the Beijing's main urban to see if it is suitable for new tourism and leisure facilities. This article simulates 949 suitable sites from the 1548 grids in the Beijing's main urban, and then further filters out 500 sites that need to be prioritized according to the population density of each main urban area to initially realize tourism at a refined scale. Quantitative site selection of leisure facilities. The research found that: ①Comparing the model simulation results with the existing tourism and leisure facilities, the accurate percentage is 95.1%, indicating the level of reliabity. ②The prediction results show that the newly added tourism and leisure facilities in the Beijing's main urban are mainly located in the densely populated areas of the Beijing's main urban. ③ The research attempts to use machine learning algorithms to plan the site selection of tourism and leisure facilities, which can optimize the overall layout and avoid the influence of subjectivity in planning on site selection. It also has certain reference value for the site selection method of other public facilities.