Training of a fuzzy classifier with polyhedral regions
説明
We discuss a fuzzy classifier with polyhedral regions. First, we generate an initial convex hull with the maximum dimension using the data, included in a class, in the general positions. Next, we modify the convex hull using one training datum at a time by the dynamic convex hull generation method. Finally, for each convex hull we define a membership function using the minimum operator and tune the slopes of the membership functions using the training data. We demonstrate the effectiveness of our method using two benchmark data sets.
収録刊行物
-
- IJCNN'01. International Joint Conference on Neural Networks. Proceedings (Cat. No.01CH37222)
-
IJCNN'01. International Joint Conference on Neural Networks. Proceedings (Cat. No.01CH37222) 2 1350-1355, 2002-11-13
IEEE