Dynamically temporal association rules mining based on SFVS
基于SFVS的时序关联规则动态发现方法

There are many disadvantages of discovering the association rules directly on the temporal series, such as high time complexity, low efficiency. So This paper considered that introduced the temporal series method based the SFVSstatistic feature vector symbolic algorithm into the association rules discovering. Used the mean and variance describing the characteristics of temporal series data as the component describing the average and the degree of divergences, that was to make the temporal series be vectors. Then discovered the association rules dynamically. The experimental results show that the association rules discovered based SFVS algorithm have a higher accuracy and reliability.