回归分析综合模型在农用地定级估价中的应用  

APPLICATION OF REGRESSIVE ANALYSIS COMPREHENSIVE MODEL IN FARMLAND GRADE-APPRAISAL

作　　者：邓旭艳[1,2] 邓良基[1,2] 张世熔[1,2] 高学松[1,2] 李宏娟[1,2] 

机构地区：[1]四川农业大学资源环境学院,四川雅安625014 [2]四川省土地信息重点实验室,四川雅安625014

出　　处：《长江流域资源与环境》2006年第A01期37-40,共4页Resources and Environment in the Yangtze Basin

基　　金：国家“863计划”《四川农业科技110》资助.

摘　　要：为了使农用地级别划分和基准地价评估真正统一，以黄金坪水电站水库淹没区为例，将淹没区农用地划分为119个定级估价单元，在Arcmap／info，SPSS软件技术平台上，用回归分析法建立水库淹没区农用地定级因素与纯收益之间关系的综合模型，划分农用地级别，然后用收益还原法计算各级别基准地价。结果为：估算纯收益与实际纯收益间绝对误差为0．05元／m^2，相对误差为13．85％，相关系数达0．86；一级地价23元／m^2，二级地价19元／m^2，三级地价15元／m。，四级地价12元／m^2。表明：回归分析综合模型在淹没区农用地定级估价中取得了较理想的效果，使定级估价数据资料直接共享，省去了专家确定权重的难关，适合水库淹没区农用地市场不发育的现状。In order to unify farmland classification and land price,this paper takes Huangjingping hydroelectric station reservoir area for example,using 119 evaluation units of agriculture land in the area. Based on the software of SPSS and Arcmap/info, by regressive analysis method, a comprehensive model of the connection between reservoir area farmland evaluation factors and net return was set up. The farmland grades are also determined. Then,it calculates land-value of every farmland level by means of income reverting. The result is that the absolute error of estimated net return and real net return is 0.05 yuan/m^2 , the relative error is 13.85%,and the correlation coefficient is 0.86. The land price of stage I is 23 yuan/m^2,19 yuan/m^2 for stage Ⅱ ,15 yuan/m^2 for stage m ,and 12 yuan/m^2 for stage Ⅳ. It shows that regressive analysis comprehensive model acquires fine reversion. The grade-appraisal data can be shared by each other. The problems that specialists must ascertain weights have been solved. The model matches the situation about sterile nature of the farmland.

关 键 词：农用地定级 农用地估价 回归分析 水库淹没区 

分 类 号：F293.3[经济管理—国民经济]