An adaptive fuzzy classification system

The problem of the data analysis and the pattern recognition, searching the relationship between the feature variables of a database and inferred results are special important. In this paper, a fuzzy classification model is established to solve the classification problem. And the objective is to propose an adaptive classification system that can be generating the fuzzy IF-THEN rules automatically and revising the confidence value dynamically. The dynamic adaptive modification algorithm is employed to modify the confidence value while that rule becomes an essential factor for classification problem. Finally, the well-known Iris and Wine databases are exploited to test the performances. Simulations demonstrate that the proposed method can provide sufficiently high classification rate even with higher feature dimension.