An Information-Theoretic Analysis of the Cost of Decentralization for Learning and Inference under Privacy Constraints

Keyphrases

• Agent Coordination 25%
• Bayesian Framework 25%
• Conditional mutual Information 25%
• Decentralization 100%
• Horizontal Federated Learning 25%
• Information Theory 25%
• Information-theoretic Analysis 100%
• Inter-agent Communication 25%
• Learning Phase 25%
• Learning Problems 25%
• Loss Cost 25%
• Multiple Agents 25%
• Performance Loss 25%
• Privacy Constraint 100%
• Supervised Learning 25%
• Vertical Federated Learning 25%

Computer Science

• Bayesian Framework 50%
• Federated Learning 100%
• Learning Phase 50%
• Learning Problem 50%
• Multiple Agents 50%
• Mutual Information 50%
• Performance Loss 50%
• Supervised Learning 50%
• Theoretic Analysis 100%