Astrophysical data mining with GPU. A case study: Genetic classification of globular clusters
Abstract
We present a multi-purpose genetic algorithm, designed and implemented with GPGPU/CUDA parallel computing technology. The model was derived from our CPU serial implementation, named GAME (Genetic Algorithm Model Experiment). It was successfully tested and validated on the detection of candidate Globular Clusters in deep, wide-field, single band HST images. The GPU version of GAME will be made available to the community by integrating it into the web application DAMEWARE (DAta Mining Web Application REsource, http://dame.dsf.unina.it/beta_info.html), a public data mining service specialized on massive astrophysical data. Since genetic algorithms are inherently parallel, the GPGPU computing paradigm leads to a speedup of a factor of 200× in the training phase with respect to the CPU based version.
Additional Information
© 2013 Elsevier B. V. Received 31 October 2012; Received in revised form 18 February 2013; Accepted 1 April 2013; Available online 21 April 2013. This work originated from a M.Sc. degree in Informatics Engineering done in the context of a collaboration among several Italian academic institutions. The hardware resources were provided by Dept. of Computing Engineering and Systems and S.Co.P.E. GRID Project infrastructure of the University Federico II of Naples. The data mining model has been designed and developed by DAME Program Collaboration. MB wishes to thank the financial support of PRIN-INAF 2010, Architecture and Tomography of Galaxy Clusters. This work has been partially funded by LINCE project of the F.A.R.O. programme jointly financed by the Compagnia di San Paolo and by the Polo delle Scienze e delle Tecnologie of the University Federico II of Napoli and it has been carried out also thanks to a hardware donation in the context of the NVIDIA Academic Partnership program. Finally, a special thanks goes to the anonymous referee for all very useful comments and suggestions.Attached Files
Submitted - 1304.0597v1.pdf
Files
| Name | Size | Download all |
|---|---|---|
|
md5:ea81eb69617e858367f4dc5f26e01c82
|
456.7 kB | Preview Download |
Additional details
- Eprint ID
- 42333
- Resolver ID
- CaltechAUTHORS:20131108-093459004
- PRIN-INAF 2010
- INAF
- Compagnia di San Paolo
- Polo delle Scienze e delle Tecnologie
- Created
-
2013-11-08Created from EPrint's datestamp field
- Updated
-
2021-11-10Created from EPrint's last_modified field