| ABSTRACT | 第8-12页 |
| 摘要 | 第13-20页 |
| Chapter 1.Introduction | 第20-29页 |
| 1.1 Background and Significance of the Research | 第20-22页 |
| 1.2 Research Objectives | 第22-23页 |
| 1.3 Solutions to Problems and their Contributions | 第23-27页 |
| 1.4 Thesis Organization | 第27-29页 |
| Chapter 2.Literature Review | 第29-50页 |
| 2.1 Related work in IDS developments | 第29-30页 |
| 2.2 Developments in k-Means/Medoids and Outlier Techniques | 第30-33页 |
| 2.3 Works related to NB and SVM Classification | 第33-41页 |
| 2.4 Recent work in Incremental SVM Classification | 第41-45页 |
| 2.5 Supervised/Unsupervised and Incremental Learning | 第45-49页 |
| 2.6 Shortcomings of the Current Researches | 第49-50页 |
| Chapter 3.k-Medoids with Naive Bayes Classification | 第50-67页 |
| 3.1 System Model and Problem Description | 第50-51页 |
| 3.2 General Description of the Solution | 第51-52页 |
| 3.3 Comparision between k-Means and k-Medoids | 第52-53页 |
| 3.4 The Proposed Hybrid Approach | 第53-59页 |
| 3.5 Experimental Results and Analysis | 第59-66页 |
| 3.5.1 Selection of Experimental Data | 第60页 |
| 3.5.2 Description of Experimental Data | 第60-61页 |
| 3.5.3 Data Pre-processing | 第61页 |
| 3.5.4 The Experimental Procedure | 第61-62页 |
| 3.5.5 Performance Evaluation | 第62-63页 |
| 3.5.6 Analysis of the Results | 第63-66页 |
| 3.6 Chapter Summary | 第66-67页 |
| Chapter 4.k-Medoids-Outlier with SVM Classification | 第67-83页 |
| 4.1 System Model and Problem Description | 第67-68页 |
| 4.2 General Description of the Solution | 第68页 |
| 4.3 The Proposed Approach | 第68-74页 |
| 4.4 Experimental Results and Analysis | 第74-82页 |
| 4.4.1 Selection of Experimental Data | 第75-76页 |
| 4.4.2 Description and Samples of Experimental Data | 第76页 |
| 4.4.3 Data Pre-processing | 第76-77页 |
| 4.4.4 The Experimental Procedure | 第77页 |
| 4.4.5 Performance Evaluation | 第77-80页 |
| 4.4.6 Analysis of the Results | 第80-82页 |
| 4.5 Chapter Summary | 第82-83页 |
| Chapter 5.Incremental Support Vector Machine with CSV-ISVM | 第83-105页 |
| 5.1 System Model and Problem Description | 第83-84页 |
| 5.2 General Description of the Solution | 第84页 |
| 5.3 Candidate Support Vectors based Incremental SVM | 第84-95页 |
| 5.3.1 The Improved Concentric Circle Method | 第86-89页 |
| 5.3.2 The Half-Partition Strategy | 第89-93页 |
| 5.3.3 CSV Selection Algorithm and CSV-ISVM Algorithm | 第93-95页 |
| 5.4 Experimental Results and Analysis | 第95-103页 |
| 5.4.1 Experimental Data | 第95-96页 |
| 5.4.2 Description and Samples of Experimental Data | 第96页 |
| 5.4.3 Data Pre-processing | 第96页 |
| 5.4.4 The Experimental Detail | 第96-97页 |
| 5.4.5 Performance Evaluation and Analysis | 第97-103页 |
| 5.5 Chapter Summary | 第103-105页 |
| Chapter 6. Concluding Remarks | 第105-107页 |
| 6.1 Conclusion | 第105-106页 |
| 6.2 Future Work | 第106-107页 |
| References | 第107-122页 |
| Published Papers | 第122-123页 |
| Acknowledgement | 第123-124页 |
| Appendix Ⅰ:Features of Kyoto 2006+ datasets | 第124-128页 |
| Appendix Ⅱ:Experimental Data Samples | 第128-136页 |
