Title

Analytical Modeling Of Data Mining Process Based On Distributed Tuple Space

Keywords

Analytical modeling; Data mining; Parallel processing; Tuple space

Abstract

The explosive growth in data volume imposes a big challenge to traditional data mining process in that data mining algorithms tend to be computationally intensive. Parallel and distributed data mining provides an attractive solution to the large scale data mining process. Message passing based parallel computing has been widely used in diverse applications such as scientific computing, ray tracing, simulation, and recently data mining. In this paper, we will present an alternative approach based on distributed tuple space that we contend provides more convenient and efficient parallel data mining framework. Then, we describe the associated analytical performance models to investigate the scalability of the proposed architecture over the cluster computing environment. We also present the experimental results for an exemplary data mining algorithm, k nearest neighbor.

Publication Date

12-1-2005

Publication Title

Proceedings of the 2005 International Conference on Parallel and Distributed Processing Techniques and Applications, PDPTA'05

Volume

3

Number of Pages

1135-1141

Document Type

Article; Proceedings Paper

Personal Identifier

scopus

Socpus ID

60749115918 (Scopus)

Source API URL

https://api.elsevier.com/content/abstract/scopus_id/60749115918

This document is currently not available here.

Share

COinS