Open AccessRough-Fuzzy Support Vector Clustering with OWA Operators
Author(s) -
Ramiro Saltos,
Richard W. Weber
Publication year - 2022
Publication title -
inteligencia artificial
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.149
H-Index - 12
eISSN - 1988-3064
pISSN - 1137-3601
DOI - 10.4114/intartif.vol25iss69pp42-56
Subject(s) - outlier , interpretability , cluster analysis , fuzzy logic , support vector machine , fuzzy clustering , kernel (algebra) , data mining , cluster (spacecraft) , operator (biology) , pattern recognition (psychology) , computer science , mathematics , artificial intelligence , algorithm , combinatorics , biochemistry , chemistry , repressor , transcription factor , gene , programming language
Rough-Fuzzy Support Vector Clustering (RFSVC) is a novel soft computing derivative of the classical Support Vector Clustering (SVC) algorithm, which has been used already in many real-world applications. RFSVC’s strengths are its ability to handle arbitrary cluster shapes, identify the number of clusters, and e?ectively detect outliers by the means of membership degrees. However, its current version uses only the closest support vector of each cluster to calculate outliers’ membership degrees, neglecting important information that remaining support vectors can contribute. We present a novel approach based on the ordered weighted average (OWA) operator that aggregates information from all cluster representatives when computing ?nal membership degrees and at the same time allows a better interpretation of the cluster structures found. Particularly, we propose the induced OWA using weights determined by the employed kernel function. The computational experiments show that our approach outperforms the current version of RFSVC as well as alternative techniques ?xing the weights of the OWA operator while maintaining the level of interpretability of membership degrees for detecting outliers.