基于矩阵的可疑金融交易增量式更新算法研究  

The research of matrix incremental updating algorithm for suspicious financial transaction

作　　者：杨涛[1] 

机构地区：[1]西安交通大学经济与金融学院,陕西西安710061

出　　处：《西北大学学报（哲学社会科学版）》2008年第4期154-157,共4页Journal of Northwest University：Philosophy and Social Sciences Edition

基　　金：国家自然科学基金项目(70771087)

摘　　要：采用基于矩阵式的增量更新算法,解决关联规则更新中没有充分利用发现结果及效率低下的问题,并通过实际数据,验证了算法的有效性及高效性,为快速识别可疑金融交易提供了有益的参考。With the increasing of financial transaction, the update database is fast, so process of discovering suspicious financial transaction is costly. Many algorithms have been developed for the maintenance of association rules. A fast updating algorithm based on matrix was proposed. The algorithm makes full use of old data mining results and greatly reduces the times of scanning the database greatly. Experiments show that MIU is efficient.

关 键 词：反洗钱 关联规则 频繁项集 增量更新算法 

分 类 号：F830[经济管理—金融学]