基于分类误差函数的变精度粗糙集模型研究

Research on Variable Precision Rough Set Model Based on Classification Error Function

由于对每个等价类的了解程度不同,其错误分类率的上限也不相同,即每一个等价类对应一个错误分类率.通过引入一个错误分类函数得到基于分类误差函数的变精度粗糙集模型;基于分类误差函数的变精度粗糙集模型随着分类误差函数的增大,其下近似算子扩大,上近似算子缩小,边界缩小.并用一个典型例子说明基于分类误差函数的变精度粗糙集模型相对于变精度粗糙集模型其精确度提高了.

The upper bound of the error classification rate of each equivalence class is different because of the different understanding of each equivalence class, that is, each equivalence class corresponds to an error classification rate. A variable precision rough set model based on a classification error function is obtained.With the increase of a classification error function, the lower approximate operator expands, the upper approximation operators shrinks, and the border narrows. With a typical example, the accuracy of the variable precision rough set model based on a classification error function is much higher.