Knowledge acquisition approach based on rough fuzzy sets

This article presents an approach of knowledge acquisition based on rough fuzzy sets, which combines features of rough sets and fuzzy sets. The continuous attributes in the decision table are fuzzified with fuzzy membership functions. The domain partition is accomplished after establishing fuzzy similarity matrix. Attributes reduction can be obtained using rough-fuzzy dependency, and then decision rules can be acquired. At last, an example is illustrated and proves the approach is effective and practical.