https://github.com/mithaystack/variantdbscan

Variant Density-Based Spatial Clustering of Applications with Noise (VariantDBSCAN)
https://github.com/mithaystack/variantdbscan

Last synced: 2 months ago
JSON representation

Variant Density-Based Spatial Clustering of Applications with Noise (VariantDBSCAN)

• Host: GitHub
• URL: https://github.com/mithaystack/variantdbscan
• Owner: MITHaystack
• Created: 2017-04-24T18:30:59.000Z (about 8 years ago)
• Default Branch: master
• Last Pushed: 2017-11-27T22:30:08.000Z (over 7 years ago)
• Last Synced: 2025-01-05T22:42:20.356Z (4 months ago)
• Language: Cuda
• Homepage:
• Size: 5.98 MB
• Stars: 1
• Watchers: 7
• Forks: 1
• Open Issues: 0
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

        
# VariantDBSCAN

## Maximizing Clustering Throughput

Project lead: Mike Gowanlock

Relevant papers: 

* [1] Gowanlock, M., Blair, D. M. & Pankratius, V. (2016) Exploiting Variant-Based Parallelism for Data Mining of Space Weather Phenomena. In Proc. of the 30th IEEE International Parallel & Distributed Processing Symposium (IPDPS 2016). pp. 760-769 DOI: 10.1109/IPDPS.2016.10 

[[http://dx.doi.org/10.1109/IPDPS.2016.10 ]](http://dx.doi.org/10.1109/IPDPS.2016.10 )

* [2] Gowanlock, M., Blair, D. M., Pankratius, V. Optimizing Parallel Clustering Throughput in Shared Memory. IEEE Transactions on Parallel and Distributed Systems DOI: 10.1109/TPDS.2017.2675421 

[[http://dx.doi.org/10.1109/TPDS.2017.2675421]](http://dx.doi.org/10.1109/TPDS.2017.2675421)

![alt text](https://github.com/MITHaystack/VariantDBSCAN/blob/master/img/VDBSCAN.png)

Figure: Relative performance gains utilizing all of the optimizations over the sequential implementation on a space weather TEC dataset in [1, 2]. Values over the black line indicate a performance improvement. The red line indicates the performance gain from index optimizations only. See the papers above.    

We acknowledge support from NSF ACI-1442997 and NASA AIST14-NNX15AG84G.