Fast learning of graph neural networks with guaranteed generalizability: One-hidden-layer case

Keyphrases

• Accelerated Gradient Method 11%
• Binary Classification 22%
• Classification via Regression 11%
• Convergence Rate 11%
• Fast Learning 100%
• Generalizability Analysis 11%
• Generalization Error 11%
• Gradient Method 11%
• Graph Complexes 11%
• Graph Neural Network 100%
• Graph Properties 11%
• Graph Structure 11%
• Hidden Layer 100%
• Learning Algorithm 44%
• Network Parameters 11%
• Neural Network Learning 11%
• Neural Network Model 22%
• Numerical Experiments 11%
• Regression Problem 11%
• Sample Complexity 11%
• Training Data 11%
• Underlying Graph 11%
• Vanilla Gradient Descent 11%

Computer Science

• Algorithm Converges 11%
• Binary Classification 22%
• Classification Problem 22%
• Convergence Rate 11%
• Generalizability 100%
• Generalization Error 11%
• Gradient Descent 22%
• Graph Neural Network 100%
• Learning Algorithm 44%
• Network Parameter 11%
• Neural Network Model 22%
• Regression Problem 11%
• Structured Data 11%
• Training Data 11%

Mathematics

• Binary Classification 22%
• Classification Problem 22%
• Convergence Rate 11%
• Generalizability 100%
• Network Model 22%
• Neural Network 100%
• Numerical Experiment 11%
• Structured Data 11%
• Tensor 11%
• Training Data 11%