Parallel processing by implication-based neuro-fuzzy systems. (English) Zbl 1057.68668
Wyrzykowski, Roman (ed.) et al., Parallel processing and applied mathematics. 4th international conference, PPAM 2001, Nałȩczów, Poland, September 9–12, 2001. Revised papers. Berlin: Springer (ISBN 3-540-43792-4). Lect. Notes Comput. Sci. 2328, 599-607 (2002).
Summary: The system composed of various implication-based neuro-fuzzy networks in one parallel structure is proposed in this paper. Different phases of data processing are distinguished, i.e. learning, testing, and problem solving. A competetive learning of the neuro-fuzzy networks is employed. This learning method refers to the first layer, which is the same in every network. The system with fuzzy parameters of membership functions is also considered. In this case, the neuro-fuzzy architectures may be viewed as fuzzy inference neural networks with fuzzy parameters, and treated analogously to fuzzy neural networks.
For the entire collection see [Zbl 0992.00055].
For the entire collection see [Zbl 0992.00055].
MSC:
| 68T05 | Learning and adaptive systems in artificial intelligence |
