excess risk bounds
Theoretical limits on the difference between a model's expected performance and the best possible performance in a given scenario, providing a framework for understanding the performance reliability of AI methods.
- Establishing Linear Surrogate Regret Bounds for Convex Smooth Losses via Convolutional Fenchel–Young Losses
- Kernel Regression in Structured Non-IID Settings: Theory and Implications for Denoising Score Learning
- Regularized least squares learning with heavy-tailed noise is minimax optimal
- Stability and Sharper Risk Bounds with Convergence Rate $\tilde{O}(1/n^2)$