Research Article
A Uniform Parallel Optimization Method for Data Mining Grid
@INPROCEEDINGS{10.1145/1577389.1577390, author={Kun Gao and Lifeng Xi}, title={A Uniform Parallel Optimization Method for Data Mining Grid}, proceedings={1st International ICST Workshop on Artificial Intelligence in Grid Computing}, publisher={ACM}, proceedings_a={AIGC}, year={2007}, month={8}, keywords={Grid Computing Distributed Computing Data Mining Knowledge Discovery in Database Performance Optimization Parallelization.Performance Design Standardization Theory.}, doi={10.1145/1577389.1577390} }- Kun Gao
Lifeng Xi
Year: 2007
A Uniform Parallel Optimization Method for Data Mining Grid
AIGC
ACM
DOI: 10.1145/1577389.1577390
Abstract
Grid is a new solution to computationally and data intensive computing problems. Since the distributed knowledge discovery process is both data and computational intensive, the Grid is a natural platform for deploying a high performance data mining service. In order to improve the performance of data mining applications, an effective method is task parallelization. Existing mechanisms of data mining parallelization are based on NOW or SMP, it is necessary to develop new parallel mechanism for grid feature. In this paper, we present a framework for high performance DDM applications in Computational Grid environments called Data Mining Grid, with the function for decomposing data mining application into subtasks and then combine those subtasks to form directed acyclic graph. This kind of parallel mechanism decomposes application according to the actual computation power of each node in dynamic Grid environment.
