Analyzing Deep PAC-Bayesian Learning with Neural Tangent Kernel: Convergence, Analytic Generalization Bound, and Efficient Hyperparameter Selection

Keyphrases

• Generalization Bounds 100%
• Neural Tangent Kernel 100%
• Kernel Convergence 100%
• Analytic Generalization 100%
• Hyperparameter Selection 100%
• PAC-Bayesian Learning 100%
• Generalization Performance 28%
• PAC-Bayes 28%
• Probabilistic Neural Network 28%
• Training Performance 14%
• Divergence 14%
• Neural Network 14%
• Tight 14%
• Machine Learning Models 14%
• Posterior Distribution 14%
• Generalization Analysis 14%
• Theoretical Generalization 14%
• Theoretical Convergence Analysis 14%
• Training Error 14%
• Kernel Ridge Regression 14%
• Prior Distribution 14%
• PAC-Bayesian 14%
• Binding Analysis 14%
• PAC-Bayesian Bounds 14%
• C-Bound 14%
• Bayes C 14%
• Proxy Measures 14%
• Gradient Descent Training 14%

Computer Science

• Leaning Parameter 100%
• Bayesian Learning 100%
• Analytic Generalization 100%
• Neural Network 42%
• Generalization Performance 28%
• Objective Function 14%
• Gradient Descent 14%
• Kernel Ridge Regression 14%
• Training Error 14%
• Machine Learning Model 14%

Mathematics

• Bayesian 100%
• Neural Network 33%
• Objective Function 11%
• Test Set 11%
• Ridge Regression 11%