| Abstract | 第5-7页 |
| 摘要 | 第8-21页 |
| 1 Introduction | 第21-29页 |
| 1.1 Background | 第21页 |
| 1.2 Research Ideas and Methods | 第21-24页 |
| 1.2.1 Research Ideas and Methods-1 | 第21-23页 |
| 1.2.2 Research Ideas and Methods-2 | 第23-24页 |
| 1.2.3 Research Ideas and Methods-3 | 第24页 |
| 1.3 Research Questions and Research Objectives | 第24-26页 |
| 1.3.1 Research Questions and Research Objectives-1 | 第24-25页 |
| 1.3.2 Research Questions and Research Objectives-2 | 第25页 |
| 1.3.3 Research Questions and Research Objectives-3 | 第25-26页 |
| 1.4 Research Innovations/Contributions | 第26-27页 |
| 1.4.1 Research Innovations/Contributions-1 | 第26页 |
| 1.4.2 Research Innovations/Contributions-2 | 第26-27页 |
| 1.4.3 Research Innovations/Contributions-3 | 第27页 |
| 1.5 The Research Content and Structure Arrangement | 第27-29页 |
| 2 Literature Review | 第29-45页 |
| 2.1 Credit Scoring | 第29-35页 |
| 2.1.1 Credit Scoring Definitions | 第29-30页 |
| 2.1.2 Five Cs Credit Scoring Concept | 第30-32页 |
| 2.1.3 The Credit Scoring Development Procedure | 第32-34页 |
| 2.1.4 Benefits of Credit Scoring | 第34-35页 |
| 2.2 Empirical Literature | 第35-45页 |
| 2.2.1 Review of Credit Scoring: Neural Network Models | 第35-40页 |
| 2.2.2 Credit Approval Data Modeling: Probabilistic Neural Network andSupport Vector Machine | 第40-41页 |
| 2.2.3 Topological Applications of MLPs and SVMs in Financial DecisionSupport Systems | 第41-45页 |
| 3 The Neural Network Credit Scoring Models | 第45-53页 |
| 3.1 Multilayer Perceptron (MLP) | 第45-46页 |
| 3.2 Radial Basis Function (RBF) | 第46-47页 |
| 3.3 Learning Vector Quantization (LVQ) | 第47-48页 |
| 3.4 Modular Neural Networks (MNN) | 第48-49页 |
| 3.5 Probabilistic Neural Network (PNN) | 第49-50页 |
| 3.6 Self Organizing Maps (SOM) | 第50-51页 |
| 3.7 Generalized Feedforward Networks (GFFNs) | 第51-52页 |
| 3.8 Jordan/Elman Network (JEN) | 第52-53页 |
| 4 Experimental Methodology | 第53-74页 |
| 4.1 Experimental Setting-1: Multiple Hybrids Neural Networks | 第53-62页 |
| 4.1.1 Data Background | 第53页 |
| 4.1.2 Experimental Protocol | 第53页 |
| 4.1.3 Data Preprocessing | 第53-55页 |
| 4.1.4 Feature Selection | 第55-56页 |
| 4.1.5 Classification | 第56-59页 |
| 4.1.6 Performance Metrics | 第59-62页 |
| 4.2 Experimental Setting-2:Credit Approval Data Modeling-Probabilistic NeuralNetwork and Support Vector Machine | 第62-67页 |
| 4.2.1 Real-world Credit Dataset | 第62-63页 |
| 4.2.2 Overview of Classification Techniques | 第63-65页 |
| 4.2.3 Evaluation Measures | 第65-66页 |
| 4.2.4 Parameter Settings | 第66-67页 |
| 4.2.5 Cross-Validation Methodology | 第67页 |
| 4.3 Experimental Setting-3:Topological Applications of MLPs and SVMs inFinancial Decision Support Systems | 第67-72页 |
| 4.3.1 Financial Decision Support Systems Dataset | 第67-69页 |
| 4.3.2 Financial Decision Support Systems Algorithms | 第69-70页 |
| 4.3.3 MLP's Activation Functions and Hidden Layers Tuning | 第70页 |
| 4.3.4 SVM Kernel Functions and Algorithm's Parameters | 第70-71页 |
| 4.3.5 Cross-Validation Methodology | 第71页 |
| 4.3.6 Performance Evaluation | 第71页 |
| 4.3.7 Roles of Credit Rating Agencies in Financial Decision SupportSystems | 第71-72页 |
| 4.4 Integration of Experimental Settings | 第72-74页 |
| 5 Experimental Results | 第74-111页 |
| 5.1 Chinese Small Business Credit Scoring: Multiple Hybrids Neural Networks | 第74-85页 |
| 5.1.1 Base Classifiers | 第74-76页 |
| 5.1.2 Neural Network Classifiers | 第76页 |
| 5.1.3 Multiple Hybrid Classifiers | 第76-81页 |
| 5.1.4 Comparisons of the Ten Best Credit Scoring Classifiers | 第81-85页 |
| 5.2 Credit Approval Data Modeling- Probabilistic Neural Network and SupportVector Machine | 第85-95页 |
| 5.2.1 Australian Credit | 第86页 |
| 5.2.2 German Credit | 第86-87页 |
| 5.2.3 Japanese Credit | 第87页 |
| 5.2.4 Chinese Credit | 第87-88页 |
| 5.2.5 PAKDD Credit | 第88页 |
| 5.2.6 Kaggle Credit | 第88-89页 |
| 5.2.7 Average Credit Scoring Performance | 第89-91页 |
| 5.2.8 Performance Validation | 第91-92页 |
| 5.2.9 Cost of Credit Scoring Errors | 第92-94页 |
| 5.2.10 Findings | 第94-95页 |
| 5.3 Topological Applications of SVMs and MLPs in Financial Decision SupportSystems | 第95-111页 |
| 5.3.1 Financial Decision Support Systems-Credit Scoring | 第95-98页 |
| 5.3.2 Financial Decision Support Systems-Bankruptcy Prediction | 第98-101页 |
| 5.3.3 Cost of Financial Decision Errors | 第101-102页 |
| 5.3.4 Measuring Average Performance | 第102-105页 |
| 5.3.5 Reliability of the Results | 第105-109页 |
| 5.3.6 Findings | 第109-111页 |
| 6 Conclusion | 第111-118页 |
| 6.1 Research Conclusion-1 | 第111-112页 |
| 6.2 Policy Implications and Future Road Maps-1 | 第112页 |
| 6.3 Research Conclusion-2 | 第112-114页 |
| 6.4 Policy Implications and Future Road Maps-2 | 第114页 |
| 6.5 Research Conclusion-3 | 第114-116页 |
| 6.6 Policy Implications and Future Road Maps-3 | 第116-118页 |
| References | 第118-128页 |
| Appendix A | 第128-137页 |
| Appendix B | 第137-140页 |
| Publications during PhD Period | 第140-141页 |
| Acknowledgement | 第141-143页 |
| Curriculum Vitae | 第143-145页 |
