基于偏微分分类数学模型的关联挖掘改进技术研究

Research on improved association mining algorithm based on partial differential classification mathematical model

为了能够使大数据关联挖掘的精度得到进一步的提高,就提出了基于偏微分分类数学模型的关联挖掘技术,创建大数据分类数据模型,得到微分方程半正定最小特征向量,之后实现稳定解的分析。本文使用渐进有理积分得到偏分分类数学模型规则集约束条件,避免在实现大数据分析过程中出现错分及漏分。此种方式能够提高大数据分类及关联挖掘的收敛性,增强抗扰动能力,具有较强的优越性。

In order to improve the precision of large data association mining, theassociation mining technology based on partial differential classification mathematical model is proposed, the large data classification data model is created, the semi positive eigenvector of the differential equation is obtained, and then the stability solution is analyzed. In this paper, we use asymptotic rational integral to get the constraint condition of rule set for mathematical models of partial classification, and avoid misclassification and missing points in the process of realizing big data analysis. This method can improve the convergence of large data classification and association mining, enhance the ability of anti disturbance, and has strong superiority.