Extreme Value Statistics of Community Detection in Complex Networks with Reduced Network Extremal Ensemble Learning (RenEEL)

Arguably, the most fundamental problem in Network Science is finding structure within a complex network. Often, this is achieved by partitioning the network’s nodes into communities in a way that maximizes an objective function. However, finding the maximizing partition is generally a computationall...

Full description

Saved in:
Bibliographic Details
Main Authors: Tania Ghosh, Royce K. P. Zia, Kevin E. Bassler
Format: Article
Language:English
Published: MDPI AG 2025-06-01
Series:Entropy
Subjects:
Online Access:https://www.mdpi.com/1099-4300/27/6/628
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849432760220909568
author Tania Ghosh
Royce K. P. Zia
Kevin E. Bassler
author_facet Tania Ghosh
Royce K. P. Zia
Kevin E. Bassler
author_sort Tania Ghosh
collection DOAJ
description Arguably, the most fundamental problem in Network Science is finding structure within a complex network. Often, this is achieved by partitioning the network’s nodes into communities in a way that maximizes an objective function. However, finding the maximizing partition is generally a computationally difficult NP-complete problem. Recently, a machine learning algorithmic scheme was introduced that uses information within a set of partitions to find a new partition that better maximizes an objective function. The scheme, known as RenEEL, uses Extremal Ensemble Learning. Starting with an ensemble of <i>K</i> partitions, it updates the ensemble by considering replacing its worst member with the best of <i>L</i> partitions found by analyzing a reduced network formed by collapsing nodes, which all the ensemble partitions agree should be grouped together, into super-nodes. The updating continues until consensus is achieved within the ensemble about what the best partition is. The original <i>K</i> ensemble partitions and each of the <i>L</i> partitions used for an update are found using a simple “base” partitioning algorithm. We perform an empirical study of how the effectiveness of RenEEL depends on the values of <i>K</i> and <i>L</i> and relate the results to the extreme value statistics of record-breaking. We find that increasing <i>K</i> is generally more effective than increasing <i>L</i> for finding the best partition.
format Article
id doaj-art-f067752cf2ff44e3bc06d65e7df3351a
institution Kabale University
issn 1099-4300
language English
publishDate 2025-06-01
publisher MDPI AG
record_format Article
series Entropy
spelling doaj-art-f067752cf2ff44e3bc06d65e7df3351a2025-08-20T03:27:17ZengMDPI AGEntropy1099-43002025-06-0127662810.3390/e27060628Extreme Value Statistics of Community Detection in Complex Networks with Reduced Network Extremal Ensemble Learning (RenEEL)Tania Ghosh0Royce K. P. Zia1Kevin E. Bassler2Department of Physics, University of Houston, Houston, TX 77204, USADepartment of Physics, University of Houston, Houston, TX 77204, USADepartment of Physics, University of Houston, Houston, TX 77204, USAArguably, the most fundamental problem in Network Science is finding structure within a complex network. Often, this is achieved by partitioning the network’s nodes into communities in a way that maximizes an objective function. However, finding the maximizing partition is generally a computationally difficult NP-complete problem. Recently, a machine learning algorithmic scheme was introduced that uses information within a set of partitions to find a new partition that better maximizes an objective function. The scheme, known as RenEEL, uses Extremal Ensemble Learning. Starting with an ensemble of <i>K</i> partitions, it updates the ensemble by considering replacing its worst member with the best of <i>L</i> partitions found by analyzing a reduced network formed by collapsing nodes, which all the ensemble partitions agree should be grouped together, into super-nodes. The updating continues until consensus is achieved within the ensemble about what the best partition is. The original <i>K</i> ensemble partitions and each of the <i>L</i> partitions used for an update are found using a simple “base” partitioning algorithm. We perform an empirical study of how the effectiveness of RenEEL depends on the values of <i>K</i> and <i>L</i> and relate the results to the extreme value statistics of record-breaking. We find that increasing <i>K</i> is generally more effective than increasing <i>L</i> for finding the best partition.https://www.mdpi.com/1099-4300/27/6/628community detectioncomplex networksmodularityextreme value statistics
spellingShingle Tania Ghosh
Royce K. P. Zia
Kevin E. Bassler
Extreme Value Statistics of Community Detection in Complex Networks with Reduced Network Extremal Ensemble Learning (RenEEL)
Entropy
community detection
complex networks
modularity
extreme value statistics
title Extreme Value Statistics of Community Detection in Complex Networks with Reduced Network Extremal Ensemble Learning (RenEEL)
title_full Extreme Value Statistics of Community Detection in Complex Networks with Reduced Network Extremal Ensemble Learning (RenEEL)
title_fullStr Extreme Value Statistics of Community Detection in Complex Networks with Reduced Network Extremal Ensemble Learning (RenEEL)
title_full_unstemmed Extreme Value Statistics of Community Detection in Complex Networks with Reduced Network Extremal Ensemble Learning (RenEEL)
title_short Extreme Value Statistics of Community Detection in Complex Networks with Reduced Network Extremal Ensemble Learning (RenEEL)
title_sort extreme value statistics of community detection in complex networks with reduced network extremal ensemble learning reneel
topic community detection
complex networks
modularity
extreme value statistics
url https://www.mdpi.com/1099-4300/27/6/628
work_keys_str_mv AT taniaghosh extremevaluestatisticsofcommunitydetectionincomplexnetworkswithreducednetworkextremalensemblelearningreneel
AT roycekpzia extremevaluestatisticsofcommunitydetectionincomplexnetworkswithreducednetworkextremalensemblelearningreneel
AT kevinebassler extremevaluestatisticsofcommunitydetectionincomplexnetworkswithreducednetworkextremalensemblelearningreneel