几类拓展粗糙集模型属性约简研究综述

A Review of Attribute Reduction of Several Extended Rough Set Models

粗糙集属性约简算法是数据预处理的有效方法,但无法处理某些结构复杂的数据.为了进一步拓宽粗糙集的应用范围,通过扩展粗糙集模型或改进属性约简算法以提高粗糙集的数据处理能力.对模糊粗糙集、覆盖粗糙集、邻域粗糙集、决策粗糙集、变精度粗糙集等几类拓展粗糙集模型的一些经典属性约简算法和最新提出的算法进行梳理和归纳后发现,现存的算法在运行效率和空间复杂度等方面限制了拓展粗糙集模型的使用范围.当前研究中拓展粗糙集模型在约简理论完善、大数据处理、特殊数据处理等三个方面的问题依然存在,因此未来应重点结合Pawlak粗糙集属性约简算法的思想、智能算法以及其他一些理论方法来研究拓展粗糙集模型属性约简理论.

Attribute reduction algorithms based on rough set are effective methods for preprocessing data, but they are not applicable for some complex data. In order to expand the scope of application of rough set further, the data processing capacity of rough set was improved by proposing a variety of extended rough set models and improving the attribute reduction algorithms. The classical attribute reduction algorithms of fuzzy rough set, covering rough set, neighborhood rough set, decision-theoretic rough set and variable precision rough set and the latest algorithms were summarized. It was found that the operation efficiency and space complexity of the existing algorithms limit the scope of application of extended rough set models by analyzing the algorithms. Finally,the problems in terms of reduction theory, big data processing, special data processing were pointed out in current researches. And the researchers should focus on integrating the idea of attribute reduction algorithms based on Pawlak rough set, intelligent algorithms and other theoretical methods to study the attribute reduction theory of extended rough set models.