一种增量时态关联规则算法

由于时态数据都有较强的时间性,即数据会随时间的变化而变化,当前已发现的某些关联规则可能不再有效。因此,我们提出了一种时态约束下的增量式关联规则挖掘算法,随着新数据的产生,增量地更新关联规则集,尽可能地只处理新数据。该算法主要是根据时态事件模型和序列模型以及Apriori原则,在快速更新算法思想基础上产生的,在实际应用中有很大的意义,能对股票数据、银行数据、超市数据和气象数据等时态数据进行分析和研究。

An Algorithm for Mining Incremental Temporal Association Rules

Temporal data are more time-sensitive.In other words,the data will change with time,and in turn some of the current association rules may no longer be valid.Therefore,we propose an algorithm for mining incremental association rules under time-board,which incrementally update association rules with the production of new data.The algorithm is generated based on mainly the state event model,the sequence model,and the Apriori principle.In practical applications,it is of great significance on the stock data,the data such as banking,the supermarket data,and the meteorological processing data.