Submodel Partitioning in Hierarchical Federated Learning: Algorithm Design and Convergence Analysis

Keyphrases

• Learning Algorithm 100%
• Convergence Analysis 100%
• Algorithm Design 100%
• Algorithm Convergence 100%
• Design Analysis 100%
• Hierarchical Federated Learning 100%
• Federated Learning 66%
• Numerical Experiments 33%
• Cell number 33%
• DEMON 33%
• Large Animal Model 33%
• Disjoint 33%
• Full Model 33%
• Global Model 33%
• Resource-constrained 33%
• Communication Cost 33%
• Architecture-centric 33%
• Convergence Behavior 33%
• Performance Efficiency 33%
• Efficiency Trade-offs 33%
• Cell Aggregation 33%
• Non-convex Cost Function 33%
• Model Partition 33%
• Storage Cost 33%
• Target Validation 33%
• Internet of Things Devices 33%
• Local Aggregation 33%
• Wide Margin 33%
• Learning Methodology 33%
• Communication Load 33%
• Global Aggregation 33%
• Star Topology 33%
• Testing Accuracy 33%
• Topology Architecture 33%

Computer Science

• Learning Algorithm 100%
• Algorithm Design 100%
• Federated Learning 100%
• Mathematical Convergence 100%
• Communication Cost 20%
• Convergence Behavior 20%
• Internet of Things Device 20%
• Star Topology 20%