GPU Accelerated Clustering Techniques

Komal D. Nistane, Shailendra W. Shende

Abstract: Clustering is the data mining technique which is used to place the data elements into related groups. With the development of technologies the size of the data is increasing rapidly. Graphics Processing Units (GPU) are the high performance parallel processors. GPU has the ability to execute different tasks independently however at the same time with the help of every single processor. This computational feature of GPU can be used to increase the speedup of the data mining algorithms. This paper shows the comparative study of the k-means and k-medoid clustering on GPU.

Keywords: k-means clustering, k-medoid clustering, graphics processing unit GPU, CUDA