基于logistic回归的区域房地产预警模型  

Modeling of Regional Real Estate Early Warning Based on Logistic Regression

作　　者：赵军[1] 赵玉玲[1] 段琪庆[1] 

机构地区：[1]济南大学土木建筑学院,山东济南250022

出　　处：《中国人口·资源与环境》2013年第S2期454-457,共4页China Population,Resources and Environment

基　　金：山东省高等学校科技计划项目"基于GIS空间统计分析的区域房地产预警模型"(编号:J10LE54);济南大学博士基金项目"区域房地产动态预警空间数据模型研究"(编号:XBS1222)

摘　　要：房地产市场运行状况的分析研究是一个非线性问题,以非线性logistic回归方法建立区域房地产预警模型,对实验区的房地产警情进行概率预测。将客观描述房地产运行情况的警情指标作为解释变量,采集济南地区1996~2011年统计数据,通过解释变量的标准化处理,获取综合预警指数,进一步运用3σ划分预警类别,作为预警模型的因变量。通过预测济南地区2012年房地产市场运行情况表明,房市过热。The situation analysis of the real estate market is a nonlinear issue. Logistic regression method is introduced to establish the regional real estate early warning model. Purpose is to achieve probability prediction of warning intelligence. Explanatory variables are police intelligence indicators,which could describe objectively the real estate market. We collect statistical data in Jinan from 1996 to2011,and process standardization on explanatory variables so as to obtain the consolidated early warning index. Furthermore,the warning categories is used as the dependent variable of the warning model,which is divided by method of 3σ. Through of predicting the real estate market situation in Jinan on 2012,the result of the real estate market is overheat.

关 键 词：房地产预警 LOGISTIC回归 概率预测 

分 类 号：F293.3[经济管理—国民经济] F224