属性受限状态下低维冗余聚类数据快速挖掘方法研究

针对传统数据挖掘方法存在挖掘精度低、速度慢、占用内存多而不适于实际应用等缺点,提出一种属性受限状态下低维冗余聚类数据挖掘方法。通过计算低维冗余聚类数据的支持度,把低维冗余聚类数据挖掘问题转变成频繁项集挖掘问题;利用支持度与可信度对关联规则产生结果进行评价,并添加属性对其进行限制,以减少无用规则的产生。通过属性位复用方法建立候选区域,产生关联规则集,对符合关联规则集的低维冗余数据进行聚类,实现对其挖掘。实验结果表明,通过所提方法对属性受限状态下低维冗余数据进行挖掘,挖掘速度快,结果可靠。

Research on fast mining method of low dimensional redundant clustering data under attribute constrained state

Aiming at the shortcomings of traditional data mining methods, such as low mining accuracy, low speed, large memory and unsuitable for practical application, a low dimensional redundant clustering data mining method under attribute constrained condition is proposed. Through calculating support low dimensional redundant clustering data, the low dimensional redundant clustering data mining problems into the frequent itemsets mining; the support and confidence of association rules generated results were evaluated, and add restrictions on the property, to reduce the useless rules. The candidate regions are created by the attribute bit multiplexing method, and the association rules set is generated, and the low dimensional redundant data conforming to the association rules are clustered. Experimental results show that the proposed method can mining low dimensional redundant data in attribute constrained state, and the mining speed is fast and the results are reliable.