PANDORA: A Parallel Dendrogram Construction Algorithm for Single Linkage Clustering on GPU
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2024
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866913808913530880 |
|---|---|
| author | Sao, Piyush Prokopenko, Andrey Lebrun-Grandié, Damien |
| author_facet | Sao, Piyush Prokopenko, Andrey Lebrun-Grandié, Damien |
| contents | This paper presents \pandora, a novel parallel algorithm for efficiently constructing dendrograms for single-linkage hierarchical clustering, including \hdbscan. Traditional dendrogram construction methods from a minimum spanning tree (MST), such as agglomerative or divisive techniques, often fail to efficiently parallelize, especially with skewed dendrograms common in real-world data.
\pandora addresses these challenges through a unique recursive tree contraction method, which simplifies the tree for initial dendrogram construction and then progressively reconstructs the complete dendrogram. This process makes \pandora asymptotically work-optimal, independent of dendrogram skewness. All steps in \pandora are fully parallel and suitable for massively threaded accelerators such as GPUs.
Our implementation is written in Kokkos, providing support for both CPUs and multi-vendor GPUs (e.g., Nvidia, AMD). The multithreaded version of \pandora is 2.2$\times$ faster than the current best-multithreaded implementation, while the GPU \pandora implementation achieved 6-20$\times$ on \amdgpu and 10-37$\times$ on \nvidiagpu speed-up over multithreaded \pandora. These advancements lead to up to a 6-fold speedup for \hdbscan on GPUs over the current best, which only offload MST construction to GPUs and perform multithreaded dendrogram construction. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2401_06089 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | PANDORA: A Parallel Dendrogram Construction Algorithm for Single Linkage Clustering on GPU Sao, Piyush Prokopenko, Andrey Lebrun-Grandié, Damien Machine Learning Distributed, Parallel, and Cluster Computing I.2.5; F.2.0; E.1 This paper presents \pandora, a novel parallel algorithm for efficiently constructing dendrograms for single-linkage hierarchical clustering, including \hdbscan. Traditional dendrogram construction methods from a minimum spanning tree (MST), such as agglomerative or divisive techniques, often fail to efficiently parallelize, especially with skewed dendrograms common in real-world data. \pandora addresses these challenges through a unique recursive tree contraction method, which simplifies the tree for initial dendrogram construction and then progressively reconstructs the complete dendrogram. This process makes \pandora asymptotically work-optimal, independent of dendrogram skewness. All steps in \pandora are fully parallel and suitable for massively threaded accelerators such as GPUs. Our implementation is written in Kokkos, providing support for both CPUs and multi-vendor GPUs (e.g., Nvidia, AMD). The multithreaded version of \pandora is 2.2$\times$ faster than the current best-multithreaded implementation, while the GPU \pandora implementation achieved 6-20$\times$ on \amdgpu and 10-37$\times$ on \nvidiagpu speed-up over multithreaded \pandora. These advancements lead to up to a 6-fold speedup for \hdbscan on GPUs over the current best, which only offload MST construction to GPUs and perform multithreaded dendrogram construction. |
| title | PANDORA: A Parallel Dendrogram Construction Algorithm for Single Linkage Clustering on GPU |
| topic | Machine Learning Distributed, Parallel, and Cluster Computing I.2.5; F.2.0; E.1 |
| url | https://arxiv.org/abs/2401.06089 |