福清市商业基准地价评估

Commercial Standard Land Price Classification Study in Fuqing City

以福清市区为研究区,应用SOM网络建立城镇基准地价评估模型,输入层为店长、店宽、标准进深、层高修正系数、建筑物成新度和年租金等6个指标,选取有代表性的65个商业样点进行网络训练,并将训练结果与租金剥离法的分类结果进行对比.结果表明:65个商业样点可分为6类,SOM网络分类预期效果很好,其与租金剥离法的吻合率达98.5%,得出应用SOM网络模型划分城镇基准地价级别具有较高实用价值的结论.

Based on artificial neural network,this paper took Fuqing City as the study area,established the evaluation model of benchmarks for urban commercial standard land price which input layer includes length,store wide,standard spatial depth,floor-high,deterioration adjustment of buildings and annual rent,etc.And we selected representative 65 commercial samples for network training,compared the results with classification results.The results showed that：65 commercial samples can be divided into six categories,the process of SOM classification was practicable,and the goodness of fit reached 98.5%,SOM model has profound practical value to evaluate urban land price.