| 人工神经网络的基准地价评估方法研究 |
| |
| 引用本文： | 刘耀林,焦利民.人工神经网络的基准地价评估方法研究[J].地球信息科学,2002,4(4):1-6. |
| |
| 作者姓名： | 刘耀林 焦利民 |
| |
| 作者单位： | 武汉大学资源与环境科学学院,武汉430079 |
| |
| 摘 要： | 探讨了人工神经网络应用于基准地价评估的可行性 ,针对基准地价评估提出了相应的人工神经网络模型和算法改进措施 ,以及在采用神经网络的前提下 ,进行土地定级与基准地价评估的新思路。实例表明 ,人工神经网络方法具有客观、准确、通用等优点。
|
| 关 键 词： | 基准地价评估 人工神经网络 BP网络 |
| 收稿时间： | 2001-12-01; |
| 修稿时间： | 2001年12月1日 |
Artificial Neural Network Based Method for Evaluating Basic Land Price |
| |
| LIU Yaolin,JIAO Limin.Artificial Neural Network Based Method for Evaluating Basic Land Price[J].Geo-information Science,2002,4(4):1-6. |
| |
| Authors: | LIU Yaolin JIAO Limin |
| |
| Institution: | School of Resource and Environmental Science,Wuhan University,430072,Wuhan, China |
| |
| Abstract: | The paper probes the application of artificial neural network in basic land price evaluation after analyzing current methods of land evaluation and properties of artificial neural network. Two artificial neural network models for land evaluation are developed, which are called as the network model I and the network model Ⅱ. The network model I is based on the relation between general value of all factors which are taken into account in evaluation work and basic land price.The network model Ⅱ is based on the relation between the value of single factor and basic land price. The improved algorithm for back propagation of two kinds of models is presented. Taking Wuhan city as an example, the models developed in this paper are tested.After comparing different models, we can make conclusions as follows: 1.The training error of neural network model I is almost the same as the testing error of exponent model, which can reflect that the relation of general acting value of factors and basic land price is the exponent relation in the test area. 2.The training error of neural network model Ⅱ is relative smaller than that of exponent model, which shows the model can represent correctly the relation of acting factor and basic land price. The model do not need to decide the weight for each factor by the people and reduce the influence to the evaluation work by the people. 3.The result shows that the models of artificial neural network applied in land evaluation are objective and accurate. |
| |
| Keywords: | evaluation of basic land price artificial neural networks(ANN) model |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
| 点击此处可从《地球信息科学》浏览原始摘要信息 |
| 点击此处可从《地球信息科学》下载免费的PDF全文 |
|
