Performance Evaluation of Regular Decomposition and Benchmark Clustering Methods
This study compares three benchmark clustering methods—mini batch k-means, DBSCAN, and spectral clustering—with regular decomposition (RD), a new method developed for large graph data. RD is first converted so that applicable to numerical data without graph structure by changing the input into a dis...
Saved in:
Main Authors: | , |
---|---|
格式: | Other NonPeerReviewed |
出版: |
Communications in Computer and Information Science
2022
|
主題: | |
在線閱讀: | https://repository.ugm.ac.id/284263/ https://link.springer.com/chapter/10.1007/978-981-19-8069-5_12 |
標簽: |
添加標簽
沒有標簽, 成為第一個標記此記錄!
|
id |
id-ugm-repo.284263 |
---|---|
record_format |
dspace |
spelling |
id-ugm-repo.2842632023-12-06T01:58:55Z https://repository.ugm.ac.id/284263/ Performance Evaluation of Regular Decomposition and Benchmark Clustering Methods Haryo, Laura Pulungan, Reza Artificial Intelligence and Image Processing This study compares three benchmark clustering methods—mini batch k-means, DBSCAN, and spectral clustering—with regular decomposition (RD), a new method developed for large graph data. RD is first converted so that applicable to numerical data without graph structure by changing the input into a distance matrix and the output into cluster labels. The results indicate that mini batch k-means has the best overall performance in terms of accuracy, time, and space consumption. RD and spectral clustering have competitive adjusted Rand index (ARI), even though their time and space consumption is considerable and can reach 2 and 30 times greater than mini batch k-means when applied to the artificial datasets. On the other hand, DBSCAN produces ARI as low as 0% in most default cases but increases up to 100% in almost all experiments of the artificial datasets after varying the parameters. DBSCAN’s accuracy, time, and space consumption, however, are still worse than mini batch k-means. Communications in Computer and Information Science 2022 Other NonPeerReviewed Haryo, Laura and Pulungan, Reza (2022) Performance Evaluation of Regular Decomposition and Benchmark Clustering Methods. Communications in Computer and Information Science. https://link.springer.com/chapter/10.1007/978-981-19-8069-5_12 10.1007/978-981-19-8069-5_12 |
institution |
Universitas Gadjah Mada |
building |
UGM Library |
continent |
Asia |
country |
Indonesia Indonesia |
content_provider |
UGM Library |
collection |
Repository Civitas UGM |
topic |
Artificial Intelligence and Image Processing |
spellingShingle |
Artificial Intelligence and Image Processing Haryo, Laura Pulungan, Reza Performance Evaluation of Regular Decomposition and Benchmark Clustering Methods |
description |
This study compares three benchmark clustering methods—mini batch k-means, DBSCAN, and spectral clustering—with regular decomposition (RD), a new method developed for large graph data. RD is first converted so that applicable to numerical data without graph structure by changing the input into a distance matrix and the output into cluster labels. The results indicate that mini batch k-means has the best overall performance in terms of accuracy, time, and space consumption. RD and spectral clustering have competitive adjusted Rand index (ARI), even though their time and space consumption is considerable and can reach 2 and 30 times greater than mini batch k-means when applied to the artificial datasets. On the other hand, DBSCAN produces ARI as low as 0% in most default cases but increases up to 100% in almost all experiments of the artificial datasets after varying the parameters. DBSCAN’s accuracy, time, and space consumption, however, are still worse than mini batch k-means. |
format |
Other NonPeerReviewed |
author |
Haryo, Laura Pulungan, Reza |
author_facet |
Haryo, Laura Pulungan, Reza |
author_sort |
Haryo, Laura |
title |
Performance Evaluation of Regular Decomposition and Benchmark Clustering Methods |
title_short |
Performance Evaluation of Regular Decomposition and Benchmark Clustering Methods |
title_full |
Performance Evaluation of Regular Decomposition and Benchmark Clustering Methods |
title_fullStr |
Performance Evaluation of Regular Decomposition and Benchmark Clustering Methods |
title_full_unstemmed |
Performance Evaluation of Regular Decomposition and Benchmark Clustering Methods |
title_sort |
performance evaluation of regular decomposition and benchmark clustering methods |
publisher |
Communications in Computer and Information Science |
publishDate |
2022 |
url |
https://repository.ugm.ac.id/284263/ https://link.springer.com/chapter/10.1007/978-981-19-8069-5_12 |
_version_ |
1784857324021612544 |