数据库中特定奇异数据优化检测的研究

The Oracle Database Specific Research of Singular Data Optimization Detect

由于数据属性会随着Oracle中心数据库中特定奇异数据的改变而发生变化,在对数据簇的属性分类时会发生动态变化,传统的挖掘方法,是通过对数据的属性分类后进行数据特征挖掘,忽略了数据本身存在的属性对挖掘效果的影响,导致挖掘误差大的问题。提出基于改进模糊c均值自适应算法的Oracle中心数据库中特定奇异数据挖掘算法。先将Oracle中心数据库中的各个数据依据其属性进行聚类,建立聚类相似系数矩阵进行划分,依据拟属可信度的概念进行归类,在利用模糊c均值算法计算Oracle中心数据库中各个数据簇中心到聚类中心的距离,建立数据隶属函数矩阵,计算加权距离平方和的最小值及有效性函数的最小值,去除较远的数据簇,实现对Oracle中心数据库中特定奇异数据的挖掘。仿真结果表明,基于改进模糊c均值自适应算法的Oracle中心数据库中特定奇异数据挖掘算法,可有效的避免在确定特定奇异数据时存在的随机性和经验性。

Based on the improved fuzzy c-means adaptive algorithm, a specific singular data mining algorithm in Oracle central database is proposed. Firstly, the data in the Oracle central database are clustered according to their attribute, partitioned according to the cluster similarity coefficient matrix, and classified based on the concept of the membership credibility. Then, the fuzzy c-means algorithm is used to calculate the distance from the center of each data cluster to the cluster center in the Oracle central database. The data membership function matrix is established to calculate the minimum value of the weighted square sum of distance and the minimum value of the effective function, the further data cluster is removed, and the specific singular data mining in the Oracle central database is realized. Simulation results show that the proposed specific singular data mining algorithm can effectively avoid the randomness and empiricism.