利用改进粒子群算法的关联规则挖掘  

Application of improved PSO algorithm in spatial association rule mining

作　　者：许栋浩 李宏伟[1] 张铁映 孟超越 樊超 

机构地区：[1]信息工程大学地理空间信息学院,郑州450001

出　　处：《测绘科学》2016年第2期168-172,139,共6页Science of Surveying and Mapping

基　　金：国家自然科学基金面上项目(40871183;41140012;41271392;41571394);国家自然科学青年基金项目(41401463);信息工程大学地理空间信息学院硕士学位论文创新与创优基金项目(XS201511)

摘　　要：针对传统的关联规则挖掘方法需要耗费大量时间来产生频繁项的问题,该文设计了一种引入负反馈机制的改进粒子群算法。该算法采用负反馈机制,在粒子位置更新前通过判断粒子是否将落入恶劣区域,较好地避免了粒子更新过程中对无用频繁项的重复计算,从而有效减少了数据库的扫描次数。通过挖掘土地覆盖类型与地形特征的空间关联关系,结果表明:所提算法不仅可以提高空间关联规则的挖掘效率,还可以发现仅具有高置信度的易被忽视关联规则。该研究结果对空间关联分析、同位模式挖掘等具有一定的参考价值。Aiming at the problem that traditional association rule mining algorithm takes a large amount of time on generating frequent item sets,this paper designed an improved particle swarm optimization（PSO）algorithm including negative feedback mechanism.By judging whether the particle would fall into bad area before updating its position,the negative feedback-based PSO algorithm avoided the double counting to useless frequent item set well throughout the iterative process,thus effectively reduced the times on scanning database.By mining the spatial association relation between land cover types and terrain features,the experimental result showed that this algorithm could not only improve the mining efficiency,but also discovered the easily-ignored association rule which only had high confidence degree.The result provided important reference value for spatial association analysis,spatial co-location mining and other research.

关 键 词：空间关联规则 粒子群 负反馈 频繁项集 适应度 

分 类 号：P208[天文地球—地图制图学与地理信息工程]