total variation distance
A measure of the difference between two probability distributions. In the context of AI, it is often used to quantify how close a learned distribution is to a target distribution, thereby assisting in model performance assessments.
- Efficient Algorithms for Robust and Partial Semi-Discrete Optimal Transport
- Nearly-Linear Time Private Hypothesis Selection with the Optimal Approximation Factor
- Optimal Best Arm Identification under Differential Privacy
- Optimal Regret of Bandits under Differential Privacy
- Product Distribution Learning with Imperfect Advice
- Restricted Spectral Gap Decomposition for Simulated Tempering Targeting Mixture Distributions