Fuzzy tree modeling based on ε-insensitive learning method
Wei Zhang, Jianqin Mao · 2011
In this paper, a new learning method tolerant to imprecision is introduced to fuzzy tree (FT) modeling method. The learning method is called ε-insensitive learning or ε learning, where, in order to fit the FT model to real data, the ε-insensitive loss function is used. FT method adaptively partitions the input space and is irrelevant to the dimension of the input space. For the consequent parameters, we use ε learning to replace the least squares estimation method which is sensitive to outliers and function influential points. Finally, numerical examples are given to demonstrate the validity of the proposed FT modeling method based on ε-insensitive learning (ε-FT).