Location Selection of Urban Elderly Care Facilities from the Perspective of Hierarchical Aging Needs

Under the background of the rapid evolution of population aging and the increasing diversity of space requirements, the traditional "facility based on facility" pension facility planning has been difficult to accurately respond to the hierarchical needs of "medical care, living care and happy care" of the elderly. This paper takes Fuzhou City, China as an example and proposes a "aging demand - spatial scene" coupled aging facility location framework: use the types of facilities related to demand—corresponding to the three demand factors of medical dependency, life circle scale, and leisure preference—as key features in the ID3 decision tree to construct a "demand-driven hierarchical decision-making model". The research takes 48,827 500-meter grids in Fuzhou City as units, integrates POI and the population data from the seventh National Census, first simulates suitable grid locations, and then takes "the population aged 65 and above > 3,000 and accounting for more than 7%" as the high-priority constraint of elderly needs to screen out the priority landing sites in the near future. The model verification results show that the fitting degree of the prediction results with 380 existing facilities reaches 95.8%, which is significantly higher than the 85% threshold. Meanwhile, the decision tree hierarchy clearly reveals the priority order of demand for "medical facilities > real estate facilities > leisure facilities", providing interpretable data evidence for the policy of integrating medical care and elderly care. The final results show that: ① The aging population concentration areas such as Jin 'an District and Minhou County are the core areas for future new facilities; ② 460 unequipped grids in the city center have been identified as urgent gaps and need to be filled with priority. ③ The peripheral counties and districts can develop tourism and elderly care based on their landscape resources to form differentiated supply. This research has expanded the application depth of machine learning in the planning of public service facilities in terms of methodology, and provided a decision-making grammar of "explicit demand - feasible space" for Fuzhou City and other rapidly aging cities in practice.