摘要
为准确反映居民对日常利用公共服务设施便利程度的真实感受,深入探讨社区建成环境对生活便利度评价的非线性影响关系。首先基于居民对生活便利度的主观评价,利用机器学习模型解析社区建成环境对生活便利度评价影响的特征重要性、阈值效应和交互作用,其次基于关键建成环境影响要素,分类居住小区的建成环境和空间分布特征,并对应提出社区建成环境生活便利度提升策略,为城市居住社区的高质量规划建设和更新调整提供科学依据。
In order to accurately reflect the real feeling about the convenience of utilizing public service facilities in residents’daily life,and deeply understand the non-linear relationship between community built environment and this evaluation of convenience of living.Firstly,this study utilized machine learning models to analyze the feature importance,threshold and interaction effects of community built environment on the evaluation of convenience of life based on residents'subjective evaluation.Secondly,this study identified three types of residential communities by clustering key impacted factors,and proposed corresponding strategies to improve built environment for these communities.This study provides a scientific reference of high quality planning and quantitative updating for urban residential communities.
作者
刘柯琚
周典
王梦莹
刘天野
杨玙珺
LIU Keju;ZHOU Dian;WANG Mengying;LIU Tianye;YANG Yujun
出处
《建筑学报》
2024年第S2期123-128,共6页
Architectural Journal
基金
国家自然科学基金项目(51778526)
陕西省自然科学基金项目(2023-JC-QN-0622)
陕西地建-西安交大土地工程与人居环境技术创新中心开放基金项目(201912131-A5)
关键词
生活便利度
15分钟社区生活圈
建成环境
机器学习
城市设计
convenience of living
15-minute community life circles
built environment
machine learning models
urban design
