声明
摘要
Abstract
Dedication
Acronyms
Contens
List of Figures
List of Tables
1 Introduction
1.1 Background
1.2 Research Motivation
1.3 Problem Definition
1.4 Research Objectives
1.5 Contributions
1.6 Research Methodology
1.6.1 Data Collection
1.6.2 Data Transformation and Data Reduction
1.6.3 Identification of the Segments With Cluster Modeling
1.6.4 Evaluation and Profiling of the Revealed Segments
1.7 Outline of the Thesis
1.8 Summary
2 Theoretical Framework
2.1 Customer Relationship Management(CRM)
2.1.1 CRM Dimension
2.1.2 Components of Customer Relationship Management
2.1.3 CRM for Customer Segmentation
2.2 Data Mining
2.2.1 Data Mining and Other Disciplines
2.2.2 Data Mining in the Banking Industry
2.3 The CRISP-DM Process Model
2.3.1 Business Understanding and Data Understanding
2.3.2 Data Preparation and Model Building
2.3.3 Model Evaluation and Deployment
2.4 Data Mining Tasks
2.4.1 Predictive and Descriptive Model
2.4.2 Dimensionality Reduction
2.5 Clustering Techniques
2.5.1 Partitioning Methods
2.5.2 Model-based Methods
2.5.3 Hierarchical Methods
2.5.4 Density-Based Clustering
2.5.5 Evaluating Clusters
2.6 Classification Techniques
2.6.1 Decision Tree
2.6.2 Types of Decision Trees
2.6.3 Association,Prediction and Sequential Patterns
2.6.4 Apriori
2.6.5 Summary
3 Credit Card Customer Segmentation
3.1 Design of the Segmentation Process
3.2 Attribute Selection
3.3 Validation and Transformation
3.4 Data Reduction Using Principal Component Analysis(PCA)
3.5 Cluster Modeling
3.5.1 Clustering With K-means
3.5.2 Clustering With the Two-Step Algorithm
3.6 Cluster Evaluation
3.6.1 Technical Evaluation of the Clustering Solution
3.6.2 Decision Tree Profiling
3.7 Summary
4 Results and Analysis
4.1 Data Description
4.2 Dimensionality Reduction Using Principal Component Analysis
4.3 The Rotated Component Matrix
4.4 Modeling Phase for Customer Segmentation
4.5 Derived Clusters
4.6 Profiling Using CHAID Decision Trees
4.7 Summary
5 Conclusion and Future Research
5.1 Conclusions
5.2 Future Research
References
Achievements
Acknowledgements