Characterization of Relevance and Irrelevance in Empirical Learning Methods based on Rough Sets and Matroid Theory
Shusaku Tsumoto, Hiroshi Tanaka · 1994
One of the most important characteristics of empirical learning methods, such as AQ, ID3(CART), C4.5 and CN2, is that they find variables which are relevant to classification. In this paper, we define relevance in em-pirical classifier as relevance of each given attribute to apparent or predictive classification, and describe this type of relevance in terms of rough sets and matroid theory. The results show that these algorithms can be viewed as the greedy algorithms searching for appar-ent classification and that their weight functions may