error bounds
Error bounds quantify the maximum expected error of a learning algorithm in relation to a given learning task or function class. They help in assessing the reliability and performance guarantees of models.
- A Gradient Guided Diffusion Framework for Chance Constrained Programming
- COALA: Numerically Stable and Efficient Framework for Context-Aware Low-Rank Approximation
- Differential Privacy on Fully Dynamic Streams
- In-context Learning of Linear Dynamical Systems with Transformers: Approximation Bounds and Depth-separation
- Kernel-based Equalized Odds: A Quantification of Accuracy-Fairness Trade-off in Fair Representation Learning
- Learning to price with resource constraints: from full information to machine-learned prices
- Metritocracy: Representative Metrics for Lite Benchmarks
- On the Optimality of the Median-of-Means Estimator under Adversarial Contamination
- Sum Estimation under Personalized Local Differential Privacy
- Understanding the Gain from Data Filtering in Multimodal Contrastive Learning