Majority-Based Classification in Distributed Environment

Distributed database system is used to store large datasets and dataset is partitioned and stored on different machines. Analysis of data in distributed system for decision support system becomes very challenging and is an emerging research area. For decision support system, various soft computing techniques are used. Data mining also provides a number of techniques for decision support system. Classification is a two-step data mining technique in which a model is developed using available datasets to predict the class label of new data. Support vector machine is a classification technique, which is based on the concept of support vectors. In this technique, after finding a classification model, class label of new record can be assigned. The trained classification model will give correct assignment if it is developed using entire dataset. But, in distributed environment, it is very difficult to bring all the data on a single machine and then develop a model for classification. Many researchers have proposed various methods for model built up in distributed system. This paper presents a majority-based classification after the development of SVM model on each machine.