Statistical Learning: the impact of feature extraction process on the dispersion map of composite data points

摘要

In this paper, we present the impact of feature extraction process on the dispersion map of composite data points. We analyze this impact in distance space using distance graphs(scatter plots). The dispersion maps of composite data points in different feature spaces are depicted and compared using scatter plots to show differences. Finally, we implement the combined feature spaces as an extension to data life cycle to test its effectiveness on dispersion map of composite data points.