一种度量数据信息不确定性的方法

A Method of Measuring Uncertainty of Data Information

对数据信息不确定性的度量是学术界和工业界十分关注的课题,目前常用的不确定性度量方法基本上是基于方差或信息熵的,在方差计算的基础上提出了一种简单的度量方法来度量随机数据的不确定性,这种度量方法基于累积方差但与传统方的基于方差的形式有所不同,和信息熵具有类似的表示意义而且计算度更快,分析结果表明该方法可以在对离散随机数据的不确定性度量应用中作为一种选择方案。此外,对定义的度量方法的性质和效果进行了讨论。

Uncertainty measurement of data information is a topic of great interest in academy and industry,the measure meth⁃ods of uncertainty most used are based on variance or entropy,on the basis of variance calculation,a simple measurement method is proposed to measure the uncertainty of random data,this measure based on cumulative variance but difference with traditional form of variance,it has similar representation meaning to information entropy and is faster to calculate.The result show that it can be an alternative in the application of uncertainty measurement to discrete random data.Further,the properties and effect of the metrics are discussed.