Construction of Fuzzy Classifiers by a Brain Storm Optimization Algorithm
Статья в журнале
The fuzzy classifier has such important advantages as an intuitive operation logic and high interpretability of the fuzzy rules base. The development of a fuzzy classifier includes three consecutive building steps: generating a fuzzy rules base, feature selection, and optimizing the parameters of membership functions. To create a rule base, clustering methods are most often used. Wrapper methods are used for feature selection, and parameter optimization is performed either by traditional optimization methods or by metaheuristic methods. In this paper, we use a metaheuristic called Brain Storm Optimization to construct a fuzzy classifier. The use of algorithms based on this metaheuristic made it possible to obtain comparable accuracy values comparable to counterparts such as D-MOFARC and FARC-HD, with a much smaller number of rules and features.
Журнал:
- Lecture Notes in Computer Science
- Springer Science and Business Media Deutschland GmbH (Томск)
- Индексируется в Scopus
Библиографическая запись: Bardamova, M. Construction of Fuzzy Classifiers by a Brain Storm Optimization Algorithm [Electronic resource] / M. Bardamova, I. Hodashinsky, M. Svetlakov // Lecture Notes in Computer Science. –2022. – Vol. 13344. – P. 391-403. – DOI: 10.1007/978-3-031-09677-8_33
Ключевые слова:
BRAIN STORM OPTIMIZATION FUZZY RULE-BASED CLASSIFIER MEMBERSHIP FUNCTION MACHINE LEARNING FUZZY CLASSIFIERИндексируется в:
- Scopus ( https://www.scopus.com/record/display.uri?eid=2-s2.0-85134314206&origin=resultslist&sort=plf-f )