An associative classifier [17] is another type of comprehensible and popularly used classifier. The associative classification rules are generated by association rule algorithms, such as the Apriori algorithm [1], but with the right hand side of the rule being fixed as the class attribute. An associative classification rule is just a IF-THEN rule. For example, {(outlook = sunny ∧ wind = false) class = play tennis} implies that the target class is play tennis if the attributes outlook is sunny and wind is false. Because associative classification rule-generating algorithms usually consider multiple variables simultaneously, a classifier formedby these rules can be less greedy than decision trees. Indeed, previous work has successfully shown the competitive accuracy performance of associative classifiers, such as CMAR [16], CPAR [28], HARMONY [24],LAC [23] and GARC [5]. Associative classification has been popularly applied in practice, such as detecting products missing from the shelf [19], maximizing customer satisfaction [14] and predicting patients' hospitalization [27].
đang được dịch, vui lòng đợi..
