论文部分内容阅读
差别矩阵方法因其易懂,易设计被广大学者接受并使用.然而,对于大型的决策表而言,计算差别矩阵时不仅费时,而且会占用大量的存储空间,使得算法的效率不高.为了尽可能减少差别矩阵的存储空间,又能用到差别矩阵的思想,提出了区分对象对集的思想,研究发现,这样设计的方法也无法适用大型数据集.为此,在区分对象对集基础上,设计一个能计算区分对象对集的个数的函数,并用该函数设计了一个启发函数,同时给出该启发函数的快速算法,利用启发函数设计了一个高效的不完备决策系统属性约简算法,新算法时间复杂度及空间复杂度分别降为O(K C U)(其中K=max{TC(xi),xi∈U}和O(U).通过实例仿真说明该算法是一高效可行的属性约简算法.
Because of its easy to understand and easy design, the discernibility matrix method is accepted and used by most scholars.However, it is not only time-consuming and computationally expensive to store the discernibility matrix in large decision tables, The idea of distinguishing object pairs is proposed, and it is found that such a design method can not be applied to large data sets.Therefore, in order to reduce the storage space of the discernibility matrix and to use the idea of discernibility matrix, , A function that can calculate the number of pairs of objects is designed, and a heuristic function is designed by the function. At the same time, a fast algorithm of the heuristic function is given. An efficient attribute reduction algorithm is designed based on the heuristic function (K = max {TC (xi), xi∈U}, and O (U) are reduced to O (KCU) by the algorithm and the new algorithm respectively.The simulation results show that this algorithm is efficient and feasible Attribute reduction algorithm.