| 摘要 | 第4-6页 |
| Abstract | 第6-8页 |
| Overview of Dissertation | 第15-17页 |
| Chapter 1:Introduction | 第17-27页 |
| 1.1- Impact of Climate Change on Disaster Risk | 第17-18页 |
| 1.2- Fires and Global Fire Risk Challenges | 第18-20页 |
| 1.3- Key Factors Affecting Fire Risks | 第20-23页 |
| 1.4- Remote Sensing Methods for Fire Risk Assessment | 第23-26页 |
| 1.5- Research Objectives | 第26-27页 |
| Chapter 2:Existing Research Base | 第27-35页 |
| 2.1- Wildfires and Climatic Parameters | 第27-29页 |
| 2.2- Fire Risk Assessment in China | 第29-32页 |
| 2.3- Fire Risk Modeling Methods | 第32-34页 |
| 2.4- Chapter Summary | 第34-35页 |
| Chapter 3:Datasets and Key Methodological Concepts | 第35-53页 |
| 3.1- MODIS Data for Wildfires Applications | 第36-38页 |
| 3.2- TRMM Data for Precipitation | 第38-40页 |
| 3.3- MODIS Data for Evapotranspiration and Potential Evapotranspiration | 第40-41页 |
| 3.4- MODIS Data for Vegetation | 第41-43页 |
| 3.5- MODIS Data for Land Use and Land Cover | 第43-44页 |
| 3.6- Machine Learning, Current Application of Artificial Intelligence | 第44-50页 |
| 3.6.1- Supervised Machine Learning | 第45-46页 |
| 3.6.2- Unsupervised Machine Learning | 第46页 |
| 3.6.3- Semi-supervised Machine Learning | 第46-47页 |
| 3.6.4- Learning from Positive and Unlabeled Data | 第47-48页 |
| 3.6.5- Support Vector Machines | 第48-50页 |
| 3.7- Machine Learning for Fire Risk Assessment | 第50-52页 |
| 3.8- Chapter Summary | 第52-53页 |
| Chapter 4:Assessing the Impact of Climatic Parameters and their Inter-Annual Seasonal Variability on Fire Activity using Time Series Satellite Products in South China | 第53-76页 |
| 4.1- Study Area and Data | 第53-55页 |
| 4.2- Methodology for Studying Relationship between Fire and Other Environmental Parameters | 第55-58页 |
| 4.3- Spatial and Temporal Variations of Fires in South China | 第58-62页 |
| 4.4- Spatial and Temporal Variations of Climatic Parameters | 第62-65页 |
| 4.5- Correlation Analysis | 第65-69页 |
| 4.6- Discussion and Conclusions | 第69-76页 |
| Chapter 5:Modeling Fire Risk in South East China Using Semi-Supervised Machine Learning Classification Algorithm | 第76-94页 |
| 5.1- Study Area and Data | 第76-78页 |
| 5.2- Methodology for Developing Semi-Supervised Machine Learning Model for Fire Risk Estimation | 第78-83页 |
| 5.2.1- Sampling Grid Files | 第79-81页 |
| 5.2.2- Transductive Bagging PU Learning Technique | 第81-82页 |
| 5.2.3- Feature Selection | 第82-83页 |
| 5.2.4- Training Supervised Binary Support Vector Machine Learning (SVM) Model with Synthetic Minority Oversampling Technique (SMOTE) | 第83页 |
| 5.2.5- Verification of SVM Model | 第83页 |
| 5.3- Model Evaluation on Test Dataset | 第83-86页 |
| 5.4- Results and Discussion from Verification Datasets | 第86-94页 |
| Chapter 6:Conclusions | 第94-98页 |
| 6.1- Conclusion | 第94-95页 |
| 6.2- Research Distinctions | 第95页 |
| 6.3- Limitations and Future Works | 第95-98页 |
| References | 第98-110页 |
| Publications during Ph.D. Program | 第110-111页 |
| List of Participated Projects and Activities during Ph.D. Program | 第111-113页 |
| Participated Projects | 第111页 |
| Poster Presentations | 第111页 |
| Conferences/ Workshops Attended | 第111-113页 |
| Acknowledgements | 第113页 |
