基于多属性决策的统计数据分级评价模型

通过对已有的专题数据分级评价模型进行分析研究,运用数理统计和决策理论,提出一种新的统计数据分级评价模型——多属性决策分级评价模型,即首先,以常用的分级数学模型作为决策方案;然后,选取分级精度、信息量、分级间隔作为方案的属性,用熵法确定各个属性的权重;最后,提出多属性决策分级评价模型,并通过建立的试验系统,进行数据实验。理论分析和实验结果表明,多属性决策分级评价模型具有很好的评价效果,为用户快速选择分级数学模型提供一种新方法。

Classification Evaluation Model of Statistic Data Based on Multiattribute Decision-making

Since there is great significance of Classification Evaluation Model of Statistic Data in thematic mapping, GIS space analysis and data mining. This paper analyzes and researches the existent classification evaluation model, then a new classification evaluation model of statistic data the classification evaluation model of multiattribute decision-making is put forward, based on mathematical statistic and decision-making theoretics, choosing the classification precision, the quantity of information and classification distance as the attribute, and working out the value of power with the method of entropy. Finally, the classification evaluation model of multiattribute decision-making is presented. The testing system is established in which a lots of data experiments are conducted. Through experiments, it proves that the new evaluation model has a great effect on a relatively scientific method selecting quickly classification model for users.