最小相关度优化PNARC算法的审计数据关联规则挖掘模型

Research on Audit Data Association Rule Mining Model with Minimal Relevance Optimized PNARC Algorithm

为解决关系国计民生重要行业事后审计的弊端,本文针对PNARC算法在审计数据关联规则挖掘时存在的置信度约束无效、挖掘精度不高等问题,提出了一种最小相关度优化PNARC算法的审计数据关联规则挖掘模型。首先对置信度进行阈值双重优化,以提高负关联规则的程度,减少不相关的关联规则,然后对最小相关度进行概率分析,降低无关规则的产生几率。仿真实验结果表明,无论在挖掘精度还是算法运行时间上,都具有比PNARC算法更优异的性能。

In order to solve the shortcomings of post-audit related to the important industries of people's livelihood and the people's livelihood, aiming at the problems of invalid confidence constraint and low mining accuracy in the mining of association rules of audit data, this paper proposes an association rule of audit data of PNARC with minimum correlation optimization Mining models. First, the confidence threshold is double-optimized to improve the degree of negative association rules and reduce the irrelevant association rules. Then the probability of the minimum correlation is analyzed to reduce the probability of generating irrelevant rules. Simulation results show that the proposed algorithm has better performance than PNARC algorithm in both the accuracy of mining and the algorithm running time.