基于ID3算法和老年人口分布的养老设施选址——以成都市中心城区为例  

Site Selection of Elderly Facilities Based on ID3 Algorithm and Elderly Population Distribution:Taking Chengdu City Centre as an Example

作　　者：汪水敏 傅红[1] Wang Shuimin;Fu Hong

机构地区：[1]四川大学建筑与环境学院建筑与环境学院

出　　处：《中外建筑》2025年第1期49-55,共7页Chinese & Overseas Architecture

基　　金：四川省科技厅重点研发项目“公园城市的韧性协同规划设计研究及示范”(20ZDYF3322)。

摘　　要：基于成都市中心城区设施POI和老年人口数据,分析了老年人口和养老设施的分布特征。采用ID3算法进行选址预测,与实际养老设施对比,重合率超过80%,显示研究具有科学可行性。通过大数据和决策树算法的支持,选址决策的科学性得以确保。结果表明:(1)初步预测,5大城区养老设施需求大,周边区域需求小;(2)通过对成都中心城区养老设施的核密度分析,发现设施分布广泛但不均,5大城区明显多于周边,城市核心和老年人口密度高的区域设施多;(3)三环内老龄化严重的5大城区养老设施选址预测的网格点位聚集,5大城区外的部分街道也需要规划养老设施。The article analyses the distribution characteristics of the elderly population and senior care facilities based on facility POI and elderly population data in central Chengdu.The ID3 algorithm is used for site selection prediction,and the overlap rate is more than 80%when compared with the actual senior care facilities,showing that the study is scientifically feasible.With the support of big data and decision tree algorithm,the scientificity of site selection decision is ensured.The results of the study show that:(1)The preliminary prediction shows that the demand for senior care facilities in the five major urban areas is large,and the demand in the surrounding areas is small.(2)From the analysis of the core density of senior care facilities in Chengdu's central city,the distribution of facilities is extensive but uneven,with the five major urban areas having significantly more facilities than the surrounding areas,and more facilities in the core of the city and areas with a high density of elderly population.(3)Overall,the grid points predicted for the siting of elderly facilities in the five major urban areas within the Third Ring Road,where aging is serious,are clustered,and some streets outside the five major urban areas also need to be planned for elderly facilities.

关 键 词：大数据 机器学习 决策树 ID3算法 养老设施 规划布局 

分 类 号：TU113.662[建筑科学—建筑理论]