基于数字化和模糊识别的一般评价理论与方法

General Evaluation Theory and Method Based on the Digital Technology and Fuzzy Pattern Recognition Theory

将数字化技术和工程模糊集理论同时引入一般评价理论,建立了高效且更加贴近评价对象客观实际的评价理论。评价工作 可应用智能技术、因特网技术、数据仓库技术和地理信息系统等数字化技术,以提高工作效率和合理性。鉴于评价对象普遍存在模糊 性,采用陈守煜提出的相对隶属度方法,用相对隶属度来描述评价对象的中介过渡,通过定义目标函数为“全体样本对于各标准模式 间的加权广义距离平方和最小”,经构造拉格朗日函数可计算出相对隶属度的值,再计算级别变量特征值,计算结果可作为各样本隶 属级别的评判指标。

This paper combines the general evaluation theory with the digital technology and t fuzzy pattern recognition theory. A new evaluation theory is presented, which is very rational and effective. The agent, Internet, data warehouse and GIS technology of digital technology times can be applied to evaluation theory. As one of the objective facts, fuzziness exists practically. By the relative membership degree method that Chen Shou-yu presented, the author describes the intermediary among the evaluation object, and constructs the objective function to minimize the sum of least squares of weighted general weighted distances between all samples and all cluster centers and a Lagrange function. Then the relative membership degree values can be obtained. Further, the features of relative state can be obtained, which can be used as the basis of evaluation.