Selecting fuzzy rules by genetic algorithm for classification problems

Ishibuchi, Hisao; Nozaki, Kengo; Yamamoto, Naohisa

doi:10.1109/fuzzy.1993.327358

Cited by 38 publications

(13 citation statements)

References 6 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…. , m. In [10] first the sum of the compatibility is calculated for each class. Then the class with the largest value of the sum is taken as the consequent class of the rule.…”

Section: B Determination Of Consequent Classmentioning

confidence: 99%

“…Then the class with the largest value of the sum is taken as the consequent class of the rule. The procedure of determining the consequent class in the conventional fuzzy rulebased classification systems in [10] is summarized in the following procedure:…”

Section: B Determination Of Consequent Classmentioning

confidence: 99%

“…Thus if a rule has a high grade of certainty, it can be used for classification even if it has only a low compatibility grade with an input pattern. The procedure of calculating the grade of certainty is the same as the one proposed in [10]: Note that the grade of certainty cannot be calculated when there are multiple classes that have the maximal value of β in Equation (7) or the maximal cost in Equation (10) depending on whether compatibility-based or cost-based determination of the consequent class is employed. In these cases the rule is not generated as the consequent class C j cannot be determined uniquely.…”

Section: Calculation Of the Grade Of Certaintymentioning

confidence: 99%

“…Genetic algorithms can also be used to generate fuzzy classification systems. For example [10] proposed a genetic algorithm that selects a small number of fuzzy if-then rules from a large number of candidate fuzzy if-then rules. Ishibuchi et al [11] proposed a Michigan-style fuzzy genetics-based machine learning method that generates fuzzy if-then rules from training patterns.…”

Section: Introductionmentioning

confidence: 99%

“…Several heuristics for improving the performance of Michigan-style fuzzy genetics-based machine learning method were proposed in [12]. In [10] and [12] the antecedent fuzzy sets are determined by genetic algorithms while the consequent part is determined from given training patterns that are covered by the antecedent part of the rule.…”

Section: Introductionmentioning

confidence: 99%

See 4 more Smart Citations

Generating Classification Rules From Numerical Data With Misclassification Cost

Nakashima¹,

Yokota²,

Schaefer³

et al. 2006

ECMS 2006 Proceedings Edited By: W. Borutzky, A. Orsoni, R. Zobel

Self Cite

View full text Add to dashboard Cite

Abstract-This paper compares the performance of various rule-based classification systems. In the classification problems in this paper it is assumed that a misclassification cost is associated with each training pattern. Thus, the task of classification is to minimize the total sum of misclassification costs rather than to maximize the classification rate. Ifthen rules are being generated from a given set of training patterns. The differences between the classification systems used in this paper are (a) whether fuzzy sets or interval sets are used in the antecedent part of ifthen rules, and (b) how the consequent part of the if-then rules is determined. In the determination of the consequent part of if-then rules we consider cost-based and compatibility-based determination . In costbased determination, the consequent class of a rule is determined so that the misclassification costs is minimal over the covered training patterns by the antecedent part of the rule. On the other hand, in compatibilitybased determination the consequent class of an if-then rule is determined from the compatibility of training patterns covered by the antecedent part of the rule. The grade of certainty of the rules in both determination types is calculated by using the compatibility of training patterns from each class. In a series of computational experiments, we examine the performance of the classification systems for three real-world pattern classification problems.

show abstract

“…. , m. In [10] first the sum of the compatibility is calculated for each class. Then the class with the largest value of the sum is taken as the consequent class of the rule.…”

Section: B Determination Of Consequent Classmentioning

confidence: 99%

Section: B Determination Of Consequent Classmentioning

confidence: 99%

Section: Calculation Of the Grade Of Certaintymentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%