A parallel ADMM-based convex clustering method
Abstract Convex clustering has received recently an increased interest as a valuable method for unsupervised learning. Unlike conventional clustering methods such as k-means, its formulation corresponds to solving a convex optimization problem and hence, alleviates initialization and local minima pr...
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
SpringerOpen
2022-11-01
|
| Series: | EURASIP Journal on Advances in Signal Processing |
| Subjects: | |
| Online Access: | https://doi.org/10.1186/s13634-022-00942-8 |
| _version_ | 1797988496980312064 |
|---|---|
| author | Lidija Fodor Dušan Jakovetić Danijela Boberić Krstićev Srđan Škrbić |
| author_facet | Lidija Fodor Dušan Jakovetić Danijela Boberić Krstićev Srđan Škrbić |
| author_sort | Lidija Fodor |
| collection | DOAJ |
| description | Abstract Convex clustering has received recently an increased interest as a valuable method for unsupervised learning. Unlike conventional clustering methods such as k-means, its formulation corresponds to solving a convex optimization problem and hence, alleviates initialization and local minima problems. However, while several algorithms have been proposed to solve convex clustering formulations, including those based on the alternating direction method of multipliers (ADMM), there is currently a limited body of work on developing scalable parallel and distributed algorithms and solvers for convex clustering. In this paper, we develop a parallel, ADMM-based method, for a modified convex clustering sum-of-norms (SON) formulation for master–worker architectures, where the data to be clustered are partitioned across a number of worker nodes, and we provide its efficient, open-source implementation (available on Parallel ADMM-based convex clustering. https://github.com/lidijaf/Parallel-ADMM-based-convex-clustering . Accessed on 10 June 2022) for high-performance computing (HPC) cluster environments. Extensive numerical evaluations on real and synthetic data sets demonstrate a high degree of scalability and efficiency of the method, when compared with existing alternative solvers for convex clustering. |
| first_indexed | 2024-04-11T08:03:45Z |
| format | Article |
| id | doaj.art-85928b9025fe4aa5b0b29b12a57362d6 |
| institution | Directory Open Access Journal |
| issn | 1687-6180 |
| language | English |
| last_indexed | 2024-04-11T08:03:45Z |
| publishDate | 2022-11-01 |
| publisher | SpringerOpen |
| record_format | Article |
| series | EURASIP Journal on Advances in Signal Processing |
| spelling | doaj.art-85928b9025fe4aa5b0b29b12a57362d62022-12-22T04:35:37ZengSpringerOpenEURASIP Journal on Advances in Signal Processing1687-61802022-11-012022113310.1186/s13634-022-00942-8A parallel ADMM-based convex clustering methodLidija Fodor0Dušan Jakovetić1Danijela Boberić Krstićev2Srđan Škrbić3Department of Mathematics and Informatics, Faculty of Sciences, University of Novi SadDepartment of Mathematics and Informatics, Faculty of Sciences, University of Novi SadDepartment of Mathematics and Informatics, Faculty of Sciences, University of Novi SadDepartment of Mathematics and Informatics, Faculty of Sciences, University of Novi SadAbstract Convex clustering has received recently an increased interest as a valuable method for unsupervised learning. Unlike conventional clustering methods such as k-means, its formulation corresponds to solving a convex optimization problem and hence, alleviates initialization and local minima problems. However, while several algorithms have been proposed to solve convex clustering formulations, including those based on the alternating direction method of multipliers (ADMM), there is currently a limited body of work on developing scalable parallel and distributed algorithms and solvers for convex clustering. In this paper, we develop a parallel, ADMM-based method, for a modified convex clustering sum-of-norms (SON) formulation for master–worker architectures, where the data to be clustered are partitioned across a number of worker nodes, and we provide its efficient, open-source implementation (available on Parallel ADMM-based convex clustering. https://github.com/lidijaf/Parallel-ADMM-based-convex-clustering . Accessed on 10 June 2022) for high-performance computing (HPC) cluster environments. Extensive numerical evaluations on real and synthetic data sets demonstrate a high degree of scalability and efficiency of the method, when compared with existing alternative solvers for convex clustering.https://doi.org/10.1186/s13634-022-00942-8Distributed optimizationADMMHigh-performance computingPerformance evaluation |
| spellingShingle | Lidija Fodor Dušan Jakovetić Danijela Boberić Krstićev Srđan Škrbić A parallel ADMM-based convex clustering method EURASIP Journal on Advances in Signal Processing Distributed optimization ADMM High-performance computing Performance evaluation |
| title | A parallel ADMM-based convex clustering method |
| title_full | A parallel ADMM-based convex clustering method |
| title_fullStr | A parallel ADMM-based convex clustering method |
| title_full_unstemmed | A parallel ADMM-based convex clustering method |
| title_short | A parallel ADMM-based convex clustering method |
| title_sort | parallel admm based convex clustering method |
| topic | Distributed optimization ADMM High-performance computing Performance evaluation |
| url | https://doi.org/10.1186/s13634-022-00942-8 |
| work_keys_str_mv | AT lidijafodor aparalleladmmbasedconvexclusteringmethod AT dusanjakovetic aparalleladmmbasedconvexclusteringmethod AT danijelaboberickrsticev aparalleladmmbasedconvexclusteringmethod AT srđanskrbic aparalleladmmbasedconvexclusteringmethod AT lidijafodor paralleladmmbasedconvexclusteringmethod AT dusanjakovetic paralleladmmbasedconvexclusteringmethod AT danijelaboberickrsticev paralleladmmbasedconvexclusteringmethod AT srđanskrbic paralleladmmbasedconvexclusteringmethod |