Extension of the Levenberg-Marquardt algorithm for the extraction of trapezoidal and general piecewise linear fuzzy rules
This paper discusses how training algorithms for determining membership functions in fuzzy rule based systems can be applied. There are several training algorithms, wbicb have been developed initially for neural networks mnd can be adapted to fumy systems. In this paper the Levenberg-Marquardt algor...
| Autor principal: | |
|---|---|
| Outros Autores: | , |
| Formato: | conferenceObject |
| Idioma: | eng |
| Publicado em: |
2013
|
| Texto completo: | http://hdl.handle.net/10400.1/2300 |
| País: | Portugal |
| Oai: | oai:sapientia.ualg.pt:10400.1/2300 |
| Resumo: | This paper discusses how training algorithms for determining membership functions in fuzzy rule based systems can be applied. There are several training algorithms, wbicb have been developed initially for neural networks mnd can be adapted to fumy systems. In this paper the Levenberg-Marquardt algorithm is introduced, allowing the determination of an optimal rukbase and converging faster tban some more classic methods (e.g. the standard Back Propagation algorithm). The class of membership funetions investigated is the trapezoidal one as it is general enough and widely used. The method can be easily extended to arbitrary piecewise linear function as well. |
|---|