基于模糊遗传算法的数据库异常数据挖掘

Data Mining Based on Fuzzy Genetic Algorithm

对大型数据库的异常数据准确挖掘是实现数据库系统的故障诊断和检测的关键技术。异常数据具有复杂性和多样性,传统方法难以对其进行准确、有效识别。为了提高异常数据挖掘性能,提出一种基于改进模糊遗传算法的大型数据库异常数据挖掘算法。构建大型数据库的异常数据信息特征模型,数据训练样本在进行遗传迭代状态下执行更新平滑,依据平方差函数值较小为原则更新簇的中心点,求得异常数据的功率谱密度函数作为特征,进行异常数据特征优选,计算异常数据流信息聚焦在多层空间模糊聚类中心,将训练集与所属的类别进行关联,得到异常数据的属性集分类和信息增益,从而提高数据的挖掘性能。仿真实验结果表明,该算法具有较高的异常数据检测和挖掘性能,挖掘识别能力优于传统模型,具有较好的应用价值。

Mining the abnormal data of the large database is the key technology to realize the fault diagnosis and detection of the database system. The abnormal data is complex and diverse, it is difficult for the traditional method to identify the abnormal data accurately and effectively. In order to improve the performance of abnormal data mining, an abnormal data mining algorithm in large databases based on improved fuzzy genetic algorithm is proposed, the characteristic model of the abnormal data information in large databases is built, data training samples are performed update smoothing in the state of genetic iteration, the cluster center is updated based on the principle that the square difference function value is smaller, then the power spectral density function of the abnormal data is obtained, and used as feature to select preference features of the abnormal data. Then the abnormal data flow information in the fuzzy clustering center is calculated, the training set is associated with the class, attribute set classification and the information gain of the abnormal data is obtained, the performance of data mining is improved. Simulation results show that the algorithm has a high performance of abnormal data detection and data mining, the mining recognition ability is better than the traditional model, and has good application value.