Fuzzy Clustering Approaches Based on Numerical Optimizations of Modified Objective Functions

Fuzzy clustering is a form of unsupervised learning that assigns the elements of a dataset into multiple clusters with varying degrees of membership rather than assigning them to a single cluster. The classical Fuzzy C-Means algorithm operates as an iterative procedure that minimizes an objective fu...

Full description

Saved in:
Bibliographic Details
Main Authors: Erind Bedalli, Shkelqim Hajrulla, Rexhep Rada, Robert Kosova
Format: Article
Language:English
Published: MDPI AG 2025-05-01
Series:Algorithms
Subjects:
Online Access:https://www.mdpi.com/1999-4893/18/6/327
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849423740300951552
author Erind Bedalli
Shkelqim Hajrulla
Rexhep Rada
Robert Kosova
author_facet Erind Bedalli
Shkelqim Hajrulla
Rexhep Rada
Robert Kosova
author_sort Erind Bedalli
collection DOAJ
description Fuzzy clustering is a form of unsupervised learning that assigns the elements of a dataset into multiple clusters with varying degrees of membership rather than assigning them to a single cluster. The classical Fuzzy C-Means algorithm operates as an iterative procedure that minimizes an objective function defined based on the weighted distance between each point and the cluster centers. The algorithm operates decently in many datasets but struggles with datasets that exhibit irregularities in overlapping shapes, densities, and sizes of clusters. Meanwhile, there is a growing demand for accurate and scalable clustering techniques, especially in high-dimensional data analysis. This research work aims to address these infirmities of the classical fuzzy clustering algorithm by applying several modification approaches on the objective function of this algorithm. These modifications include several regularization terms aiming to make the algorithm more robust in specific types of datasets. The optimization of the modified objective functions is handled based on several numerical methods: gradient descent, root mean square propagation (RMSprop), and adaptive mean estimation (Adam). These methods are implemented in a Python environment, and extensive experimental studies are conducted, following carefully the steps of dataset selection, algorithm implementation, hyper-parameter tuning, picking the evaluation metrics, and analyzing the results. A comparison of the features of these algorithms on various datasets is carefully summarized.
format Article
id doaj-art-ed7dbda62afe46d481270d153c311eee
institution Kabale University
issn 1999-4893
language English
publishDate 2025-05-01
publisher MDPI AG
record_format Article
series Algorithms
spelling doaj-art-ed7dbda62afe46d481270d153c311eee2025-08-20T03:30:29ZengMDPI AGAlgorithms1999-48932025-05-0118632710.3390/a18060327Fuzzy Clustering Approaches Based on Numerical Optimizations of Modified Objective FunctionsErind Bedalli0Shkelqim Hajrulla1Rexhep Rada2Robert Kosova3Department of Computer Engineering, Epoka University, 1039 Tirana, AlbaniaDepartment of Computer Engineering, Epoka University, 1039 Tirana, AlbaniaDepartment of Informatics, University of Elbasan “Aleksandër Xhuvani”, 3001 Elbasan, AlbaniaDepartment of Mathematics, University “Aleksander Moisiu” Durres, 2001 Durres, AlbaniaFuzzy clustering is a form of unsupervised learning that assigns the elements of a dataset into multiple clusters with varying degrees of membership rather than assigning them to a single cluster. The classical Fuzzy C-Means algorithm operates as an iterative procedure that minimizes an objective function defined based on the weighted distance between each point and the cluster centers. The algorithm operates decently in many datasets but struggles with datasets that exhibit irregularities in overlapping shapes, densities, and sizes of clusters. Meanwhile, there is a growing demand for accurate and scalable clustering techniques, especially in high-dimensional data analysis. This research work aims to address these infirmities of the classical fuzzy clustering algorithm by applying several modification approaches on the objective function of this algorithm. These modifications include several regularization terms aiming to make the algorithm more robust in specific types of datasets. The optimization of the modified objective functions is handled based on several numerical methods: gradient descent, root mean square propagation (RMSprop), and adaptive mean estimation (Adam). These methods are implemented in a Python environment, and extensive experimental studies are conducted, following carefully the steps of dataset selection, algorithm implementation, hyper-parameter tuning, picking the evaluation metrics, and analyzing the results. A comparison of the features of these algorithms on various datasets is carefully summarized.https://www.mdpi.com/1999-4893/18/6/327fuzzy clusteringobjective functionregularization termsnumerical methodsgradient descentroot mean square propagation
spellingShingle Erind Bedalli
Shkelqim Hajrulla
Rexhep Rada
Robert Kosova
Fuzzy Clustering Approaches Based on Numerical Optimizations of Modified Objective Functions
Algorithms
fuzzy clustering
objective function
regularization terms
numerical methods
gradient descent
root mean square propagation
title Fuzzy Clustering Approaches Based on Numerical Optimizations of Modified Objective Functions
title_full Fuzzy Clustering Approaches Based on Numerical Optimizations of Modified Objective Functions
title_fullStr Fuzzy Clustering Approaches Based on Numerical Optimizations of Modified Objective Functions
title_full_unstemmed Fuzzy Clustering Approaches Based on Numerical Optimizations of Modified Objective Functions
title_short Fuzzy Clustering Approaches Based on Numerical Optimizations of Modified Objective Functions
title_sort fuzzy clustering approaches based on numerical optimizations of modified objective functions
topic fuzzy clustering
objective function
regularization terms
numerical methods
gradient descent
root mean square propagation
url https://www.mdpi.com/1999-4893/18/6/327
work_keys_str_mv AT erindbedalli fuzzyclusteringapproachesbasedonnumericaloptimizationsofmodifiedobjectivefunctions
AT shkelqimhajrulla fuzzyclusteringapproachesbasedonnumericaloptimizationsofmodifiedobjectivefunctions
AT rexheprada fuzzyclusteringapproachesbasedonnumericaloptimizationsofmodifiedobjectivefunctions
AT robertkosova fuzzyclusteringapproachesbasedonnumericaloptimizationsofmodifiedobjectivefunctions