Neuro-fuzzy system with learning tolerant to imprecision

In this paper, a new teaming method tolerant to imprecision is introduced and used in neuro-fuzzy modeling. This method can be called e-insensitive teaming, where in order to fit the fuzzy model to real data, a weighted e-insensitive loss function is used. The proposed method makes it possible to exclude an intrinsic inconsistency of neuro-fuzzy modeling, where zero-tolerance teaming is used to obtain a fuzzy model tolerant to imprecision. The e-insensitive teaming leads to a model with the minimal Vapnik-Chervonenkis dimension (complexity), which results in improving generalization ability of this system and its robustness to outliers. Finally, numerical examples are given to demonstrate the validity of the introduced method. (C) 2002 Elsevier B.V. All rights reserved.