Improvement of Pre-processing Capacity of Support Vector Clustering using Neural Network Kernel Function for Stream Data Classification

Abstract

Pre-processing of data before generation of pattern or classification is major steps. In the phase of pre-processing reduces the noise level of data using different technique of data mining. In current research trend support vector clustering is used for efficient data processing for noise reduction and pattern generation. Support vector clustering is new paradigm of data mining tools. It combined with supervised learning and unsupervised learning. for the success story behind support vector clustering technique is kernel function. The better selection of kernel function produces better result in terms of noise reduction and classification. In this paper proposed an improved support vector clustering method using neural network kernel function for stream data classification. The neural network function work as data optimizer and data selector in support vector clustering.