Clustering explanation based on multi-hyperrectangle

Abstract Clustering plays a crucial role in data mining and pattern recognition, but the interpretation of clustering results is often challenging. Existing interpretation methods usually lack an intuitive and accurate description of irregular shapes and high dimensional datas. This paper proposes a...

Full description

Saved in:
Bibliographic Details
Main Authors: Tao Zeng, Caiming Zhong, Tiejun Pan
Format: Article
Language:English
Published: Nature Portfolio 2024-12-01
Series:Scientific Reports
Online Access:https://doi.org/10.1038/s41598-024-81141-3
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1846137103847522304
author Tao Zeng
Caiming Zhong
Tiejun Pan
author_facet Tao Zeng
Caiming Zhong
Tiejun Pan
author_sort Tao Zeng
collection DOAJ
description Abstract Clustering plays a crucial role in data mining and pattern recognition, but the interpretation of clustering results is often challenging. Existing interpretation methods usually lack an intuitive and accurate description of irregular shapes and high dimensional datas. This paper proposes a novel clustering explanation method based on a Multi-HyperRectangle(MHR), for extracting post hoc explanations of clustering results. MHR first generates initial hyperrectangles to cover each cluster, and then these hyper-rectangles are gradually merged until the optimal shape is obtained to fit the cluster. The advantage of this method is that it recognizes the shape of irregular clusters and finds the optimal number of hyper-rectangles based on the hierarchical tree structure, which discovers structural relationships between rectangles. Furthermore, we propose a refinement method to improve the tightness of the hyperrectangles, resulting in more precise and comprehensible explanations. Experimental results demonstrate that MHR significantly outperforms existing methods in both the tightness and accuracy of cluster interpretation, highlighting its effectiveness and innovation in addressing the challenges of clustering interpretation.
format Article
id doaj-art-09d2c237080b4e4b80549427e4ef22f3
institution Kabale University
issn 2045-2322
language English
publishDate 2024-12-01
publisher Nature Portfolio
record_format Article
series Scientific Reports
spelling doaj-art-09d2c237080b4e4b80549427e4ef22f32024-12-08T12:27:28ZengNature PortfolioScientific Reports2045-23222024-12-0114111810.1038/s41598-024-81141-3Clustering explanation based on multi-hyperrectangleTao Zeng0Caiming Zhong1Tiejun Pan2College of Science and Technology, Ningbo UniversityCollege of Science and Technology, Ningbo UniversityCollege of Science and Technology, Ningbo UniversityAbstract Clustering plays a crucial role in data mining and pattern recognition, but the interpretation of clustering results is often challenging. Existing interpretation methods usually lack an intuitive and accurate description of irregular shapes and high dimensional datas. This paper proposes a novel clustering explanation method based on a Multi-HyperRectangle(MHR), for extracting post hoc explanations of clustering results. MHR first generates initial hyperrectangles to cover each cluster, and then these hyper-rectangles are gradually merged until the optimal shape is obtained to fit the cluster. The advantage of this method is that it recognizes the shape of irregular clusters and finds the optimal number of hyper-rectangles based on the hierarchical tree structure, which discovers structural relationships between rectangles. Furthermore, we propose a refinement method to improve the tightness of the hyperrectangles, resulting in more precise and comprehensible explanations. Experimental results demonstrate that MHR significantly outperforms existing methods in both the tightness and accuracy of cluster interpretation, highlighting its effectiveness and innovation in addressing the challenges of clustering interpretation.https://doi.org/10.1038/s41598-024-81141-3
spellingShingle Tao Zeng
Caiming Zhong
Tiejun Pan
Clustering explanation based on multi-hyperrectangle
Scientific Reports
title Clustering explanation based on multi-hyperrectangle
title_full Clustering explanation based on multi-hyperrectangle
title_fullStr Clustering explanation based on multi-hyperrectangle
title_full_unstemmed Clustering explanation based on multi-hyperrectangle
title_short Clustering explanation based on multi-hyperrectangle
title_sort clustering explanation based on multi hyperrectangle
url https://doi.org/10.1038/s41598-024-81141-3
work_keys_str_mv AT taozeng clusteringexplanationbasedonmultihyperrectangle
AT caimingzhong clusteringexplanationbasedonmultihyperrectangle
AT tiejunpan clusteringexplanationbasedonmultihyperrectangle