| Abstract | 第5-6页 |
| 摘要 | 第7-14页 |
| List of Symbols | 第14-15页 |
| List of Acronyms | 第15-20页 |
| Chapter 1 Introduction | 第20-30页 |
| 1.1 Overview | 第20-26页 |
| 1.1.1 Distributed machine learning | 第22-23页 |
| 1.1.2 Distributed cooperative adaptation | 第23-25页 |
| 1.1.3 Distributed multi-agent optimization | 第25-26页 |
| 1.2 Contributions and organization | 第26-30页 |
| Chapter 2 Preliminaries | 第30-36页 |
| 2.1 Algebraic graph theory | 第30页 |
| 2.2 Distributed average consensus | 第30-31页 |
| 2.3 Feedforward neural network with random weights | 第31-33页 |
| 2.4 RBF neural networks | 第33-36页 |
| Chapter 3 A Zero-Gradient-Sum Algorithm for Distributed Cooperative Learning Basedon FNNRW | 第36-56页 |
| 3.1 Introduction | 第36-37页 |
| 3.2 Problem formulation | 第37-40页 |
| 3.3 ZGS-based distributed optimization model | 第40-45页 |
| 3.4 Distributed cooperative learning for FNNRW | 第45-48页 |
| 3.5 Simulations | 第48-56页 |
| 3.5.1 Test case 1: approximation of the ‘Sin C’ function with noise | 第48-52页 |
| 3.5.2 Test case 2: classification of the MNIST dataset | 第52-56页 |
| Chapter 4 Distributed Cooperative Learning for FNNRW Based on Event-TriggeredCommunication | 第56-74页 |
| 4.1 Introduction | 第56-58页 |
| 4.2 Problem formulation | 第58-60页 |
| 4.3 Distributed learning for FNNRW with event-triggered communication | 第60-64页 |
| 4.3.1 ZGS-based distributed learning for FNNRW | 第60-62页 |
| 4.3.2 Distributed learning for FNNRW with event-triggered communication | 第62-64页 |
| 4.4 Convergence rate | 第64-70页 |
| 4.5 Simulations | 第70-74页 |
| 4.5.1 Test case1: approximation of the ‘Sin C’ function with noise | 第70-72页 |
| 4.5.2 Test case 2: classification of the MNIST dataset | 第72-74页 |
| Chapter 5 Distributed Cooperative Learning for Output Feedback RBF Neural Net-works | 第74-96页 |
| 5.1 Introduction | 第74-75页 |
| 5.2 Preliminaries | 第75-77页 |
| 5.3 DCL from adaptive neural network output control | 第77-88页 |
| 5.3.1 Problem description and controller design | 第77-79页 |
| 5.3.2 DCL scheme | 第79-80页 |
| 5.3.3 Closed-loop stability and neural network learning capability | 第80-88页 |
| 5.4 Learning control using experiences | 第88-90页 |
| 5.5 Simulations | 第90-96页 |
| Chapter 6 Population-Based Distributed Optimization over Peer-to-Peer Networks | 第96-118页 |
| 6.1 Introduction | 第96-98页 |
| 6.2 Population-based distributed optimization framework | 第98-99页 |
| 6.3 D-PSO-ON algorithm | 第99-109页 |
| 6.3.1 Consensus search | 第102-105页 |
| 6.3.2 Consensus evaluation | 第105-106页 |
| 6.3.3 PSO cooperative evolution | 第106-108页 |
| 6.3.4 Terminating with local silencing rule | 第108-109页 |
| 6.4 Simulations | 第109-115页 |
| 6.4.1 Test case 1: a small undirected/directed networks | 第110-114页 |
| 6.4.2 Test case 2: a large undirected network | 第114-115页 |
| 6.5 Discussions | 第115-118页 |
| Chapter 7 Conclusions | 第118-122页 |
| References | 第122-134页 |
| Acknowledgments | 第134-136页 |
| Curriculum Vitae | 第136-138页 |
