A GPGPU Approach to Accelerate Ant Swarm Optimization Rough Reducts (ASORR) Algorithm
Reducts can be used to discern all discernible objects from the original information system. In order to find a reducts, a such applications of rough set uses a discernibility matrix. Ant Swarm Optimization Rough Reducts (ASORR) algorithm is used in rough reducts calculation for identifying signific...
Saved in:
| Main Authors: | , , , |
|---|---|
| 格式: | Conference or Workshop Item |
| 語言: | English |
| 出版: |
2012
|
| 主題: | |
| 在線閱讀: | http://eprints.utem.edu.my/id/eprint/6466/1/citee2012_submission_91.pdf http://eprints.utem.edu.my/id/eprint/6466/ |
| 標簽: |
添加標簽
沒有標簽, 成為第一個標記此記錄!
|
| 機構: | Universiti Teknikal Malaysia Melaka |
| 語言: | English |
| 總結: | Reducts can be used to discern all discernible objects from the original information system. In order to find a reducts, a such applications of rough set uses a discernibility matrix. Ant Swarm Optimization Rough Reducts (ASORR) algorithm is used in rough reducts calculation for identifying significant attribute set optimally. For the complex matrix calculation in a single cpu, it will take a long computing time to build the discernibility matrix whereas the execution time of an algorithm is needed to be considered. This paper proposed an parallel approach to accelerate the execution time of ASORR algorithm which is utilizing GPGPU that supports high speed parallel computing. It will be implemented with CUDA from NVIDIA. The experiment results indicate that parallel ASORR achieve the acceleration on GPGPU. |
|---|