论文部分内容阅读
本文针对房产特征价格估价模型特征变量量化和函数拟合不准确的问题,提出了基于GIS和PCA建立住宅房产特征价格模型的思路:利用GIS管理分析功能辅助特征变量量化,避免了量化的主观随意性,提高量化精度;利用PCA方法提取房产特征变量量化值的主成分,然后进行拟合分析,从而解决函数拟合中变量相关性问题,提高了函数拟合准确性;以构建的模型为基础开发了房产估价决策支持系统软件,最后利用该软件系统对赣州房产进行估价实践,表明该模型能有效提高房产估价精度和可信度。
In this paper, aiming at the problem of quantification of feature variables and inaccuracy of function fitting, a method to set up the characteristic price model of residential real estate based on GIS and PCA is put forward. Using GIS management analysis function to assist quantification of characteristic variables avoids the subjective and arbitrary quantification And improve the precision of quantification. The PCA method is used to extract the principal components of the quantized values of house property variables, and then the fitting analysis is conducted to solve the problem of variable dependency in function fitting and improve the accuracy of function fitting. The model is based on the constructed model The real estate appraisal decision support system software was developed. At last, the software system was used to evaluate the real estate in Ganzhou, which shows that the model can effectively improve the appraisal accuracy and credibility of real estate.