Customer Churn Analysis Model in Manufacturing Industry

D.S. Liu; Chun Hua Ju

doi:10.4028/www.scientific.net/AMR.69-70.675

Paper Titles

Research of Quickly Dry Equipments about Chinese Herbal Medicine
p.655

Multi-Motor Synchronization of Multi-Wire Saw Based on Model-Free Adaptive Control
p.660

A Tension Control Approach to the Multi-wire Saw Based on DSP
p.665

A Fast Algebraic Grid Generation Algorithm Base on an Improved Kamaugh Map
p.670

Customer Churn Analysis Model in Manufacturing Industry
p.675

Calculating the Maximum Throughput in the Context of Workflow Execution
p.680

Study of Active Vibration Control for Flexible Beam’s Vibration
p.685

On Pneumatic Installation Technology of Modal Testing Suspension System
p.690

An Extended CPM for Resource-Constrained Project Scheduling
p.695

HomeAdvanced Materials ResearchAdvanced Materials Research Vols. 69-70Customer Churn Analysis Model in Manufacturing...

Customer Churn Analysis Model in Manufacturing Industry

Abstract:

To address the problem of customer churn in CRM in manufacturing industry, this paper proposes a prediction model based on Support Vector Machine (SVM). Considering the large-scale and imbalanced churn data, principal component analysis (PCA) is adopted to reduce dimensions and eliminate redundant information, which makes the sample space for SVM more compact and reasonable. An improved SVM is used to predict customer churn. Firstly, PCA is adopted to process 17 dimensional feature vectors of customer churn data, and then the application in manufacturing industry verifies that this model based on both PCA and SVM performs better than the model based on SVM only and other traditional models.