【摘要】 针对对海量数据库中的大数据进行优化挖掘,可以提高数据特征的提取和检测能力.传统方法采用模糊C均值聚类的数据挖掘算法,当数据在层次聚类过程中空间特征的相似度差异性较小时,数据挖掘的准确度不高.提出一种基于粒子群混沌差分训练对模糊C均值聚类算法进行改进,建立数据挖掘优化模型.首先提出了数据聚类据挖掘模型的总体构架,采用非线性时间序列分析方法进行数据信息流拟合,对数据信息流进行高阶累积量特征提取,采用粒子群混沌差分训练实现模糊C均值聚类算法改进.以改进的模糊聚类算法对提取的高阶累积量特征进行聚类分析,以分析结果为依据对数据挖掘模型进行优化.仿真结果表明,该数据挖掘模型能有效实现海量数据的优化聚类和特征提取,数据挖掘的精度较高,性能较好,避免挖掘过程陷入局部收敛.更多还原

【Abstract】 To optimize the massive big data in the database mining,can improve the data feature extraction and detection ability.Traditional method using fuzzy c-means clustering data mining algorithm,when the data in the process of hierarchical clustering space characteristics of similarity difference is small,the accuracy of data mining is not high.In this paper,a chaos particle swarm optimization difference training to improve the fuzzy c-means clustering algorithm,establish the optimization model for data mining.First puts forward the data clustering according to the overall architecture of the mining model,data structure analysis,nonlinear time series analysis method is adopted to improve the flow of information data fitting,higher-order cumulant features of data streams are extracted,using particle swarm chaos difference fuzzy c-means clustering algorithm to improve the training implementation.With the improved fuzzy clustering algorithm to extract the higher-order cumulant features for clustering analysis,based on the results of the analysis of the data mining model optimization.The simulation results show that the data mining model can effectively realize the optimization of huge amounts of data clustering and feature extraction,data mining of high precision,good performance,avoid digging into local convergence.更多还原