A Unified Perspective on Conformal Prediction and Wasserstein Distributionally Robust Optimization for Uncertainty Quantification
The episode discusses a paper unifying Conformal Prediction (CP) and Distributionally Robust Optimization (DRO) for uncertainty quantification. Hosts explain that while both methods aim for high coverage, they achieve it differently—one through level inflation and the other through value-space correction. This provides a robust toolkit for building verifiable, trustworthy AI systems.