摘要
基于某市农村危房改造管理信息系统和实地调研数据,采用数据标准化、基于聚类分析等数据挖掘方法,对农户的房屋建筑年代、家庭人口数量以及现住房建筑面积进行了离群点检测。运用聚类分析,选取农户家庭人口、现住房面积等能区分农户类型的数据,将其分为3类农户。结合实地调查农户危房改造意愿与原因,以及问卷调研数据,提出了通过信息增益理论及C4.5决策树建立农户危房改造意愿与农户特征间的关联关系的方法。研究发现家庭人口数量多及现住房面积较小的农户更愿意改造危房。研究结果为危房改造工作提供方法和思路上的借鉴。
Based on a city’s rural dilapidated house retrofitting management information system and field survey data,data mining methods such as data standardization and cluster analysis were used to detect the out-of-group points of the housing construction era,the number of households,and the current housing construction area.Using cluster analysis,the data of farmer households and current housing area can be divided into three types of farmers.Combined with the on-the-spot investigation of farmers’willingness and reason for the renovation of dilapidated houses,and the questionnaire survey data,a method is proposed to establish the relationship between farmers’willingness to change and the characteristics of farmers through data gain theory and C4.5 decision tree.The study found that farmers with large family size and small housing area are more willing to retrofit dilapidated houses.It can assist the staff to understand the difficulties of farmers and provide reference for methods and ideas for the renovation of dilapidated houses.
作者
张舸
周志鹏
周浩
ZHANG Ge;ZHOU Zhi-peng;ZHOU Hao(School of Civil and Resource Engineering,University of Science and Technology Beijing,Beijing 100083,China;Institute for Urban Governance and Sustainable Development,Tsinghua University,Beijing 100084,China;Key Laboratory of Eco Planning&Green Building,Ministry of Education(Tsinghua University),Beijing 100084,China)
出处
《建筑节能（中英文）》
CAS
2022年第12期106-110,118,共6页
Building Energy Efficiency
基金
国家重点研发计划资助项目“基于全生命周期碳减排的建筑运行能效和健康性能提升研究”(2018YFE0106100)。
关键词
农村危房改造
数据挖掘
聚类分析
离群点
信息增益
C4.5决策树
住房面积
rural dilapidated housing rehabilitation
data mining
cluster analysis
outlier detection
information divergence
C4.5 decision tree
housing area
