Rough set approach for distributed decision tree and attribute reduction in the disseminated environment

Abstract

Attribute reduction is a necessitated step for the disseminated environment in regard to classification and prediction of the data. Traditional approaches were not efficient for optimal attribute reducts. Current techniques are quiet time consuming and less accuracy. Rough set approach is a mathematical technique to handle attribute reducts through data dependencies and structural methods. This paper discusses a novel algorithm for optimal attribute deduct and also increases the accuracy in predicted results and the distributed decision tree classification techniques was made use of to implement the same in the disseminated environment. Proposed algorithm for Construction of distributed decision trees with rough sets increases the accuracy and also reduces the attributes on the time of the massive data sets handling. © 2011 Springer-Verlag.