Gradient Scheduling with Global Momentum for Asynchronous Federated Learning in Edge Environment

Keyphrases

• Edge Habitat 100%
• Gradient Scheduling 100%
• Non-IID 100%
• Asynchronous Federated Learning 100%
• Edge Nodes 75%
• Asynchronous Training 50%
• Convergence Rate 25%
• Optimization Algorithm 25%
• Performance Degradation 25%
• Performance Gain 25%
• Training Process 25%
• Data Distribution 25%
• Training Methods 25%
• Accuracy Improvement 25%
• Novel Paradigm 25%
• Federated Learning 25%
• Computational Capability 25%
• Machine Learning Models 25%
• Training Stability 25%
• MNIST 25%
• Biased Gradients 25%
• Popular 25%
• CIFAR10 Dataset 25%

Computer Science

• Federated Learning 100%
• Performance Gain 50%
• Performance Degradation 50%
• Training Process 50%
• Data Distribution 50%
• Convergence Rate 50%
• Optimization Algorithm 50%
• Machine Learning 50%
• Learning System 50%