wasserstein distance
Wasserstein distance is a metric used to measure the difference between probability distributions, particularly useful in generative models and optimal transport problems, as it accounts for how distributions can be transformed into each other.
- Bootstrap Your Uncertainty: Adaptive Robust Classification Driven by Optimal-Transport
- Composite Flow Matching for Reinforcement Learning with Shifted-Dynamics Data
- Differentiable Generalized Sliced Wasserstein Plans
- KAIROS: Scalable Model-Agnostic Data Valuation
- MARS: A Malignity-Aware Backdoor Defense in Federated Learning
- Pareto Optimal Risk-Agnostic Distributional Bandits with Heavy-Tail Rewards
- Self-Evolving Pseudo-Rehearsal for Catastrophic Forgetting with Task Similarity in LLMs
- Stratify or Die: Rethinking Data Splits in Image Segmentation