A novel parallel clustering algorithm based on artificial immune network using nVidia CUDA framework

Authors:
Ruiyi Luo;Qian Yi
Affiliations:
Image Processing and Pattern Recognition Laboratory, Beijing Normal University, Beijing, China;Image Processing and Pattern Recognition Laboratory, Beijing Normal University, Beijing, China
Venue:
HCII'11 Proceedings of the 14th international conference on Human-computer interaction: design and development approaches - Volume Part I
Year:
2011

Citing 5
Cited 0

Data mining: concepts and techniques

Data mining: concepts and techniques
An Evolutionary Immune Network for Data Clustering

SBRN '00 Proceedings of the VI Brazilian Symposium on Neural Networks (SBRN'00)
Design of parallel algorithms for fractal video compression

International Journal of Computer Mathematics - Distributed Algorithms in Science and Engineering
A Parallel Immune Algorithm Based on Fine-Grained Model with GPU-Acceleration

ICICIC '09 Proceedings of the 2009 Fourth International Conference on Innovative Computing, Information and Control
A comparative study of real-valued negative selection to statistical anomaly detection techniques

ICARIS'05 Proceedings of the 4th international conference on Artificial Immune Systems

Quantified Score

Hi-index	0.00

Visualization

Abstract

In this paper, a novel parallel data clustering algorithm based on artificial immune network aiNet is proposed to improve its efficiency. In consideration of the restrictions of GPU, we carefully designed the data structure, algorithm flow and memory allocation strategy of the parallel algorithm and realized it using NVIDIA's CUDA framework. During the implementation, in order to fully explore its implicit parallelism, we allocated threads on GPU that represent the network cells of aiNet, and modified this algorithm to let those thread operations parallel during the clustering process. We calculated the affinity parallel, combined the random numbers with the local search algorithm to select the first n cell parallel, and did the network suppression parallel. Experimental results show that certain speedup can be obtained by using the proposed method.