论文部分内容阅读
属性约简是粗糙集理论一个重要的研究问题.在粗糙集理论上,利用粒计算的思想构建了粒矩阵,提出并定义了粒矩阵相与运算,建立了基于粒矩阵的知识粒化方法,并且给出了粒矩阵属性约简的启发式算法.采用粒矩阵进行属性约简选择最小属性集,跳出了传统属性约简的先求解属性核,再求解最优属性集的方法.理论分析表明了新的算法是可靠有效的,给粒计算属性约简提供一个新的思路,为进一步研究粒计算提供可行的方法.
Attribute reduction is an important research issue in rough set theory.In the rough set theory, the granular matrix is constructed by using the idea of granular computing, the granular matrix is proposed and defined, and the granulation method based on granular matrix is established. And a heuristic algorithm for attribute reduction of the granular matrix is given.It adopts the granular matrix to reduce the attribute to select the minimum attribute set and jump out of the solution kernel of the traditional attribute reduction and then the method of solving the optimal attribute set.Theoretical analysis shows The new algorithm is reliable and effective, and provides a new idea for the granular computing attribute reduction, which provides a feasible method for further study of granular computing.