An Improved Algorithm of Fuzzy Association Rules and Its Application in IDS
Wenguo Wang · Computer Technology and Development · 2007
Discovery of association rule is an important problem in database mining,but it is merely used to handle the discrete data.To partition continuous quantitative attribute is handled by using fuzzy partition in order to solve the problem of sharpening boundary,which provides a smooth transition of data partition.On IDS the requirements of training data are very low.In the paper,an improved algorithm using Hashing tables on mining fuzzy association rules is proposed,and equivalence classes are introduced to search frequent itemsets quickly.With this algorithm the usual practice of repeatedly database scanning can be avoided.Its efficiency is showed with a typical use on intrusion detection system(IDS)from network datasets.