Distributed Data Mining On Grid Environment

Abstract

Data mining tasks considered a very complex business problem. In this research, we study the enhancement in the speedup of executing data mining tasks on a grid environment. Experiments were performed by running two main data mining algorithms Classification and Clustering algorithms, and one of the data sampling methods for classification task which is Cross Validation. These tasks were executed on large dataset. Gird environment was prepared by installing GridGain framework on the experimental machines which were connected by a LAN. Experimental results show significant enhancement in the speedup when executing data mining tasks on a grid of computing nodes.