| 초록 |
In this paper we propose a FNLPLS (Fuzzy nonlinear PLS) model, which combines strong points of Takagi-Sugeno-Kang (TSK) fuzzy and LWR model. This model is a kind of parametric regression model, and provides a powerful tool for modeling complex nonlinear systems. It enhances the interpretability of NLPLS and is strengthened by its learning algorithm comprising global and local learning. Moreover, it permits us to adjust its parameter according to the usage in terms of global fitting and local interpretation. |
