一种基于免疫算法的空间关联规则挖掘方法

针对海量数据空间关联规则挖掘的不足,提出了一种基于免疫算法的空间关联规则挖掘方法。算法充分利用了免疫识别、免疫记忆及克隆选择特性,把要挖掘的空间关联规则作为抗原,候选项目集作为抗体,把挖掘的关联规则存入记忆库,加快了关联规则的挖掘速度。以杆塔故障的空间要素的关联关系为例,验证了算法的有效性。

A New Spatial Association Rules Mining Method Based on Immune Algorithms

On the basis of analyzing the now-generally-used spatial association rules algorithm,aiming at the shortage of the very large database spatial association rules mining,a spatial association rules mining algorithm based on immune algorithms is proposed.This algorithm makes use of the immune recognition mechanism,immune memory characters and clonal selection characters.In the process of spatial association mining,spatial association rules are regarded as the antigens,candidate itemsets are looked upon as the antibodies.The spatial association rules are stored in memory,and speed of mining spatial association rules is accelerated.We take the incidence relation of special data of pole and tower fault as an example,to verify the algorithm.Experiment results show that the proposed algorithm is effective.The algorithm is able to be more quickly and efficiently search in the whole global,and extremely be used for the mining spatial association rules to very large database.