Bossley, K.M., Brown, M. and Harris, C.J.,
Neurofuzzy Adaptive Modelling and Construction of Nonlinear Dynamical Processes
Irwin, G.R., Warwick, K. and Hunt, K.J. (eds.)
At Neural Network Applications in Control.
Full text not available from this repository.
The identification of nonlinear dynamical processes has become an important task in many different areas of research. The formulation of such models is inherently a very difficult task. Neurofuzzy modelling has recently been proposed to help tackle this idetification problem, where neural networks and fuzzy logic are combined, providing fuzzy systems to which thorough mathematical analysis can be applied. Fundamental to system identification is the principle of parsimony, where the best model is the one with simplest acceptable structure. This coupled with the curse of dimensionality has lead to the development of efficient off-line parsimonious neurofuzzy construction algorithms. This chapter discusses a range of neurofuzzy algorithms that automatically construct parsimonious models. In this discussion different construction algorithms and alternative (non-lattice based) neurofuzzy models are addressed.
Conference or Workshop Item
|Venue - Dates:
||Neural Network Applications in Control, 1995-01-01
||Southampton Wireless Group
||04 May 1999
||18 Apr 2017 00:24
|Further Information:||Google Scholar|
Actions (login required)