AutoSF: Adaptive Distributed Model Training in Dynamic Edge Computing

Keyphrases

• Edge Computing 100%
• Aggregation Structure 100%
• Aggregation Rate 100%
• Distributed Model Training 100%
• Dynamic Edge 100%
• Edge Habitat 60%
• Convergence Rate 20%
• Numerical Experiments 20%
• Heuristic Approach 20%
• Near-optimal 20%
• Environment Change 20%
• Data Distribution 20%
• Evaluation Results 20%
• Training Performance 20%
• Machine Learning Approach 20%
• Distributed Learning 20%
• Benchmark Approach 20%
• Edge Resources 20%
• AutoML 20%
• AI Models 20%
• Main Idea 20%
• Self-developed 20%
• Edge Devices 20%
• Model Aggregation 20%
• Training Accuracy 20%
• Optimal Aggregation 20%
• Global Loss Function 20%

Computer Science

• Distributed Model 100%
• Edge Computing 100%
• Machine Learning Approach 33%
• Testbed 33%
• Evaluation Result 33%
• Distributed Learning 33%
• Automated Machine Learning 33%
• Convergence Speed 33%
• Data Distribution 33%
• Artificial Intelligence 33%