Selection of Comparative Cases in Land Appraisal Based on CloudModel and Gray Relevancy Theory

摘要

There is a traditional approach named market comparative method, which is used to appraise land price by putting a focus on comparing cases. A review of the developing history of this method highlights some persistent challenges. In this paper, this existing approach is extended through coupling the cloud model, a data-mining technique, with gray relevancy theory. This approach allows the construction of quantitative measurement according to qualitative concept (attribute), simulating human cognizing process. This novel method admits hierarchical describing and exploration of the relationship and proximity between district factors and individual factors. By 1-D cloud generator, we obtain several rules in terms of linguistic atom which confirm the number of overlap cloud. Each rule corresponds to a rule's, rear that achieves uncertainty ratiocination. In order to prove applicability, of this method, it is applied to the selection of comparative cases in land appraisal of Wuhan City. Experimental results show most cases can be correctly discriminated and the better comparative cases are acceptable. Compared with other approaches, this method has better performance in land appraisal.