PAC-Bayes Generalisation Bounds for Dynamical Systems Including Stable RNNs

Mathematics

• Error Bound 100%
• Dynamical System 100%
• Neural Network 100%
• Bayesian 50%
• Nonlinear 50%
• Illustrative Example 50%
• Time Series Data 50%
• Loss Function 50%
• Data Distribution 50%
• Mixing Coefficient 50%
• Linear Dynamical Systems 50%
• Essential Supremum 50%
• Discrete Time 50%

Keyphrases

• Generalization Bounds 100%
• Generalization Gap 100%
• Lipschitz 50%
• Essential Supremum 50%
• Data Time Series 50%
• Mixing Coefficient 50%
• PAC-Bayes Bound 50%