high-dimensional settings
High-dimensional settings refer to scenarios where datasets have a large number of features compared to the number of samples. This challenges learning algorithms due to potential overfitting and dimensionality-related issues.
- BayeSQP: Bayesian Optimization through Sequential Quadratic Programming
- Causal Discovery over Clusters of Variables in Markovian Systems
- Differentiable Generalized Sliced Wasserstein Plans
- Identifying Macro Causal Effects in C-DMGs over DMGs
- Non-Asymptotic Analysis Of Data Augmentation For Precision Matrix Estimation
- On the Closed-Form of Flow Matching: Generalization Does Not Arise from Target Stochasticity
- On the Closed-Form of Flow Matching: Generalization Does Not Arise from Target Stochasticity
- QuanDA: Quantile-Based Discriminant Analysis for High-Dimensional Imbalanced Classification
- The Unseen Threat: Residual Knowledge in Machine Unlearning under Perturbed Samples