摘要
在AprioriTid算法的基础上提出了两点改进:一是利用压缩的候选项集代替数据库D,减少了数据量;二是根据优化的频繁k-1项集L′k-1来生成候选项目集Ck,避免了不必要的组合.实验证明,改进算法在缩小数据库规模方面是行之有效的.
The article gives two improvements based on AprioriTid Algorthms. Firstly, it reduces data by using the compressed candidate itemset instead of the database D; Secondly, it avoids making the unnecessary combination. This method makes use of the optimized frequent itemset to create candidate itemset. The validity of the improved Algorithms in reducing database's size is proved by experiment.
出处
《山东师范大学学报(自然科学版)》
CAS
2005年第4期20-22,共3页
Journal of Shandong Normal University(Natural Science)
基金
山东省自然科学基金重大项目(Z2004G02)