The research on the data mining technology in the active demand management - Details

Author：

Indexed by：

EI Scopus

Abstract：

The　traditional　K-Means　algorithm　is　sensitive　to　outliers,　outliers　traction　and　easy　off-center,　and　overlap　of　classes　can　not　very　well　show　their　classification.　This　paper　introduces　a　variant　of　the　probability　distribution　theory,　K-Means　clustering　algorithm　-　Gaussian　mixture　model　to　part　of　the　customer　data　randomly　selected　of　Volkswagen　dealer　in　a　Beijing　office　in　2008,　for　example,　and　carry　out　empirical　study　based　on　the　improved　clustering　algorithm　model.　The　results　showed　that:　data　mining　clustering　algorithm　in　active　demand　management　and　market　segmentation　has　important　significance.　©　2011　IEEE.

Keyword：

K-means clustering Statistics Computation theory Gaussian distribution Information services Data mining

Author Community：

[ 1 ] [Xuemei, Chen]Beijing Institute of Technology, School of Mechanical and Vehicular Engineering, Beijing 100081, China
[ 2 ] [Xuemei, Chen]Jiangsu University Automobile Key Laboratory, Jiangsu University, Zhejiang, 212013, China
[ 3 ] [Li, Gao]Beijing Institute of Technology, School of Mechanical and Vehicular Engineering, Beijing 100081, China
[ 4 ] [Xi, Wang]Beijing Institute of Technology, School of Mechanical and Vehicular Engineering, Beijing 100081, China
[ 5 ] [Zhonghua, Wei]Beijing University of Technology, Traffic Engineering Key Lab. of Beijing, Beijing, China
[ 6 ] [Zhenhua, Zhang]Beijing Institute of Technology, School of Mechanical and Vehicular Engineering, Beijing 100081, China
[ 7 ] [Zhigao, Liao]School of Mechanical and Vehicular, General Station of Quality Inspection for Special Engineering of Civil Aviation 4, Beijing 100007, China