Fast discovery of high fuzzy utility itemsets

This work presents an efficient approach for deriving itemsets with high fuzzy utility values from quantitative data. Each item in a transaction has its own profit and quantity, and the total fuzzy utility of it is considered. We also design a useful strategy to prune unpromising fuzzy candidate itemsets, thus making the mining process efficient. Through a series of experimental evaluations, the results show the proposed approach could perform well in fuzzy utility mining.