Knowledge-Based Neurocomputing: A Fuzzy Logic Approach (Google eBook)
Springer Science & Business Media, Jan 17, 2009 - Computers - 100 pages
The technologies of fuzzy systems and fuzzy controllers, in particular, have been applied with great success to numerous real world applications. The number of entries in the INSPEC database with keywords fuzzy mod- ing and fuzzy control dated 1969 2007is 1,541 and 9,728, respectively. In spite of the evident progress reported in terms of concepts, algorithmic - velopments and engineering practice, there are still a number of challenging and highly relevant problems. Unfortunately, the existing publications are rather silent when it comes to reporting comprehensive solutions to them. The two challenges become particularly apparent and have been triggeredby the growing complexity of the applications. The ?rst evident challenge we are faced with is the curse of dimensionality. Rule based systems and fuzzy rule based systems are quite a?ected by this phenomenon especially when tackling problems of high dimensionality. The second one concerns a way of constructing fuzzy models which are accurate yet highly interpretable. The author of the monographhas focused on these two vital problems and o?eredan interesting, originaland practically relevantinsight into their so- tions.Whendealingwithfuzzymodeling, thebookfocusesonabroadclassof Takagi Sugeno Kang (TS) fuzzy models a highly legitimate choice given a wealthof literature onthese constructs and a greatdeal of their applications. Furthermorethe TS fuzzy models havebeen asubject ofnumerousanalytical studies which have resulted in a series of interesting ?ndings. This situation stands in a deep contrast with the most studies carried out in the realm of fuzzy control where analytical methods are not very common."
What people are saying - Write a review
We haven't found any reviews in the usual places.
AB language activation function ANN functioning apply approach Artiﬁcial Intelligence artiﬁcial neural networks atom binary strings classiﬁcation classiﬁcation problem Conf connectionism Corollary Counter Module deﬁned deleted denote diﬀerent diﬃcult digit easy to verify eﬀect equals equivalent FARB examples Extracting rules FARB–ANN equivalence feedforward ANN ﬁnite-state ﬁrst First-Order RNN formal language fuzzy logic fuzzy rules Fuzzy Sets hidden layer hidden neurons IEEE IEEE Trans If-part iﬀ implies initial knowledge input string input-output ISBN knowledge embedded Knowledge Extraction Knowledge-Based Design Knowledge-Based Neurocomputing Kolman larger than 1/2 Logistic functions machine learning mapping Margaliot modeled Neural Computation neuro-fuzzy neuron output parameters Proc recurrent neural networks representation RNN for recognizing second-order RNN sequence si(i si(t sigmoid functions simpliﬁcation procedure speciﬁc springerlink.com Step STUDFUZZ 234 substring suitable support vector machines symbolic techniques Then-part trained ANN transform values variables x3 is st yields zero