Genetic Fuzzy Systems

Cordón, Óscar; Herrera, Francisco; Hoffmann, Frank; Magdalena, Luis

doi:10.1142/4177

Cited by 615 publications

(97 citation statements)

References 0 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…For each input variable we assume here an associated term set of N L possible values, represented by uniformly distributed triangular fuzzy sets with a linguistic meaning, resulting in a descriptive 20 FRBCS. Moreover, we use fuzzy rules of the so-called disjunctive normal form (DNF), where each variable is allowed to take as value multiple linguistic labels from its associated term set, joined by the OR disjunctive operator, instead of only one.…”

Section: Basic Conceptsmentioning

confidence: 99%

“…However, traditional deterministic learning methods cannot appropriately handle highly-dimensional feature spaces, a fact that has limited so far the use of FRBCSs in multispectral data [16][17][18] . In the direction of efficiently determining the structure of FRBCSs for complex classification tasks, the enhanced search capabilities of Genetic Algorithms 19 (GAs) have been extensively used in the derivation of FRBCSs (and fuzzy rule-based systems (FRBSs) in general), giving rise to the field of genetic FRBCSs 20 (GFRBCSs). Feature selection mechanisms can easily be employed in GFRBSs, leading to compact fuzzy rule bases, thus increasing the inherent interpretability properties of FRBSs.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Reducing the Complexity of Genetic Fuzzy Classifiers in Highly-Dimensional Classification Problems

Stavrakoudis¹,

Galidaki²,

Gitas³

et al. 2012

IJCIS

View full text Add to dashboard Cite

This paper introduces the Fast Iterative Rule-based Linguistic Classifier (FaIRLiC), a Genetic Fuzzy Rule-Based Classification System (GFRBCS) which targets at reducing the structural complexity of the resulting rule base, as well as its learning algorithm's computational requirements, especially when dealing with high-dimensional feature spaces. The proposed methodology follows the principles of the iterative rule learning (IRL) approach, whereby a rule extraction algorithm (REA) is invoked in an iterative fashion, producing one fuzzy rule at a time. The REA is performed in two successive steps: the first one selects the relevant features of the currently extracted rule, whereas the second one decides the antecedent part of the fuzzy rule, using the previously selected subset of features. The performance of the classifier is finally optimized through a genetic tuning post-processing stage. Comparative results in a hyperspectral remote sensing classification as well as in 12 real-world classification datasets indicate the effectiveness of the proposed methodology in generating high-performing and compact fuzzy rule-based classifiers, even for very high-dimensional feature spaces.

show abstract

Section: Basic Conceptsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Reducing the Complexity of Genetic Fuzzy Classifiers in Highly-Dimensional Classification Problems

Stavrakoudis¹,

Galidaki²,

Gitas³

et al. 2012

IJCIS

View full text Add to dashboard Cite

show abstract

“…Thus we could not always find good rule sets. We will be able to find better rule sets with higher classification performance by directly searching for rule sets (i.e., combinations of fuzzy rules) as in Pittsburgh-style fuzzy GBML algorithms (e.g., see Cordon et al (2001)). Such a fuzzy GBML algorithm, however, usually requires long CPU time and large memory storage for finding good rule sets for high-dimensional problems.…”

Section: Genetic Algorithm-based Rule Selectionmentioning

confidence: 99%

“…On the other hand, iterative algorithms often lead to higher classification performance than Michigan-style algorithms where each fuzzy rule is represented by a string and a population corresponds to a fuzzy rule set. For further discussions on these three classes of fuzzy GBML algorithms, see Cordon et al (2001). The aim of this paper is to compare several heuristic rule selection criteria used for fuzzy rule extraction from numerical data.…”

Section: Introductionmentioning

confidence: 99%

Comparison of Heuristic Criteria for Fuzzy Rule Selection in Classification Problems

Ishibuchi

Yamamoto

2004

Fuzzy Optimization and Decision Making

View full text Add to dashboard Cite

Abstract. This paper compares heuristic criteria used for extracting a pre-specified number of fuzzy classification rules from numerical data. We examine the performance of each heuristic criterion through computational experiments on well-known test problems. Experimental results show that better results are obtained from composite criteria of confidence and support measures than their individual use. It is also shown that genetic algorithm-based rule selection can improve the classification ability of extracted fuzzy rules by searching for good rule combinations. This observation suggests the importance of taking into account the combinatorial effect of fuzzy rules (i.e., the interaction among them).

show abstract

“…This is because the use of evolutionary computation can enhance several abilities of fuzzy systems, such as the generalization ability for unseen and uncertain data sets, the interpretability for users, and the applicability to realworld problems. A series of special issues on EFSs and GFSs 7,8,9,10,11,12,13,14 and a web bibliography compilation 15 clearly show that this research field is continuously growing and breaking in new research topics: novel representation schemes 16 , interpretability of fuzzy systems 17 , scalability issues 18 , subgroup discovery 19 , imbalanced datasets 20 , etc. This special issue includes the recent novel contributions to pattern classification, regression, association rule mining, and real-world applications.…”

mentioning

confidence: 99%

Special Issue on Evolutionary Fuzzy Systems

Alcalá¹,

Nojima²,

Ishibuchi³

et al. 2012

IJCIS

Self Cite

View full text Add to dashboard Cite

During this two decades, evolutionary computation has been successfully used for the design of fuzzy systems under the name of Evolutionary Fuzzy Systems (EFSs) or Genetic Fuzzy Systems (GFSs) 1,2,3,4,5,6 . This is because the use of evolutionary computation can enhance several abilities of fuzzy systems, such as the generalization ability for unseen and uncertain data sets, the interpretability for users, and the applicability to realworld problems. A series of special issues on EFSs and GFSs 7,8,9,10,11,12,13,14 and a web bibliography compilation 15 clearly show that this research field is continuously growing and breaking in new research topics: novel representation schemes 16 , interpretability of fuzzy systems 17 , scalability issues 18 , subgroup discovery 19 , imbalanced datasets 20 , etc. This special issue includes the recent novel contributions to pattern classification, regression, association rule mining, and real-world applications.The first four papers are related to pattern classification problems. Two of them develop novel classifiers different from usual rule-based ones. The others focus on data sets with different aspects from standard benchmark data sets.The paper "An efficient inductive genetic learning algorithm for fuzzy relational rules" by A. González et al. proposes a genetic fuzzy rule learning algorithm for the design of a fuzzy relational rule model. The use of fuzzy relational rules enhances the knowledge representation ability of fuzzy models. The experimental study on pattern classification problems clearly shows that the proposed method can modify the generalization ability of fuzzy rulebased classifiers.The paper "A study on the use of multiobjective genetic algorithms for classifier selection in FURIAbased fuzzy multiclassifiers" by K. Trawiński et al. presents an ensemble fuzzy classifier design by evolutionary multiobjective algorithms. FURIA is used to generate component classifiers. NSGA-II is used to optimize the combination of the component classifiers with respect to two of four objective functions: the training accuracy, the number of component classifiers in an ensemble, and two diversity measures. The effects of different objective functions are examined through the experiments.

show abstract

Genetic Fuzzy Systems

Cited by 615 publications

References 0 publications

Reducing the Complexity of Genetic Fuzzy Classifiers in Highly-Dimensional Classification Problems

Reducing the Complexity of Genetic Fuzzy Classifiers in Highly-Dimensional Classification Problems

Comparison of Heuristic Criteria for Fuzzy Rule Selection in Classification Problems

Special Issue on Evolutionary Fuzzy Systems

Contact Info

Product

Resources

About