Fuzzy C-Means with Improved Chebyshev Distance for Multi-Labelled Data

Fuzzy C-Means (FCM) is one of the most well-known clustering algorithms, nevertheless its performance has been limited by the utilization of Euclidean as its distance metric.Even though there exist studies that applied FCM with other distance metrics such as Manhattan, Minkowski and Chebyshev, its p...

Full description

Saved in:
Bibliographic Details
Main Authors: Mousa, Aseel, Yusof, Yuhanis
Format: Article
Language:English
Published: Medwell Publishing 2018
Subjects:
Online Access:http://repo.uum.edu.my/24433/1/JEAS%2013%202%202018%20353-360.pdf
http://repo.uum.edu.my/24433/
https://www.medwelljournals.com/abstract/?doi=jeasci.2018.353.360
Tags: Add Tag
No Tags, Be the first to tag this record!
Institution: Universiti Utara Malaysia
Language: English
Description
Summary:Fuzzy C-Means (FCM) is one of the most well-known clustering algorithms, nevertheless its performance has been limited by the utilization of Euclidean as its distance metric.Even though there exist studies that applied FCM with other distance metrics such as Manhattan, Minkowski and Chebyshev, its performance can still be argued particularly on multi-label data.Various applications rely on data points that can be grouped into more than one class and this includes protein function classification and image annotation.This study proposes the employment of FCM that is implement using an improved Chebyshev distance metric.The proposed work eliminates correlation in data points and improve performance of clustering.The results show that the proposed FCM improves the performance of clustering as it produces minimum objective function value and with less iteration count. Such a result indicates that FCM with improved distance metric contributes in producing better clusters.