Achieving Globally Superlinear Convergence for Distributed Optimization with Adaptive Newton Method

Keyphrases

• Distributed Optimization 100%
• Newton Algorithm 100%
• Modified Newton Method 100%
• Superlinear Convergence 100%
• Consensus Methods 40%
• Newton's Method 20%
• Existing Algorithms 20%
• Numerical Experiments 20%
• Compress 20%
• Hessian 20%
• Peer-to-peer Networks 20%
• Local Computation 20%
• Low-rank Approximation 20%
• Communication Efficiency 20%
• Global Rate 20%
• Decision Vector 20%
• Linear Convergence 20%
• Quadratic Convergence Rate 20%
• Set Consensus 20%
• Asymptotic Consensus 20%
• Superlinear Convergence Rate 20%

Computer Science

• Newton's Method 100%
• Distributed Optimization 100%
• Newton Algorithm 100%
• Convergence Rate 40%
• Optimization Problem 20%
• Objective Function 20%
• Rank Approximation 20%
• Peer to Peer Networks 20%
• Decision Vector 20%

Mathematics

• Newton's Method 100%
• Convergence Rate 66%
• Consensus Method 66%
• Asymptotics 33%
• Numerical Experiment 33%
• Objective Function 33%
• Low-Rank Approximation 33%
• Finite Time 33%
• Linear Convergence 33%