A comparative review on nondeterministic sets for association rule mining

Nowadays decision making based on data mining mostly deals with imprecise environment. Managing various uncertainties is one of the main challenging areas in decision support systems. The aim of this paper is to compare the relationship among four paramount uncertain sets namely, soft, grey, rough and fuzzy sets. The origin of these vague names is investigated and how they could be combined to make effective usage is shown. A systematic consideration is accomplished with respect to data mining approaches. The analysis demonstrates that these uncertain sets provide different but overlapping approaches for uncertainty representation and reasonable consolidation of them in rule mining could lead to more appropriate results.