An Ensemble Method based on Particle of Swarm for the Reduction of Noise, Outlier and Core Point

The majority voting and accurate prediction of classification algorithm in data mining are challenging task for data classification. For the improvement of data classification used different classifier along with another classifier in a manner of ensemble process. Ensemble process increase the classification ratio of classification algorithm, now such par diagram of classification algorithm is called ensemble classifier. Ensemble learning is a technique to improve the performance and accuracy of classification and predication of machine learning algorithm. Many researchers proposed a model for ensemble classifier for merging a different classification algorithm, but the performance of ensemble algorithm suffered from problem of outlier, noise and core point problem of data from features selection process. In this paper we combined core, outlier and noise data (COB) for features selection process for ensemble model. The process of best feature selection with appropriate classifier used particle of swarm optimization. Empirical results with UCI data set prediction on Ecoil and glass dataset indicate that the proposed COB model optimization algorithm can help to improve accuracy and classification.