Fuzzy Association Rules Mining Algorithm FMFFI Based on Bidirectional Search Technique

Abstract

Association rules is one of the important studies on data mining, while, the study of quantitative association rules mining is lacking. This paper proposes a fuzzy association rules mining algorithm FMFFI (Fast Mining Fuzzy Frequent Item sets) based on bidirectional search. This algorithm uses FCM clustering technique to map quantitative data sets into fuzzy data sets, and uses the bidirectional search method search from the high-dimension to low-dimension and low-dimension to high-dimension, when search fuzzy frequent item sets to reduce search time and improve the data mining efficiency.