primal-dual algorithm
Primal-dual algorithms are optimization techniques that solve problems by considering both the primal and dual formulations of an optimization problem, often used in scenarios involving constraints like in structured prediction.
- Composition and Alignment of Diffusion Models using Constrained Learning
- Constrained Entropic Unlearning: A Primal-Dual Framework for Large Language Models
- Near-Optimal Sample Complexity for Online Constrained MDPs
- Semi-infinite Nonconvex Constrained Min-Max Optimization
- Structured Reinforcement Learning for Combinatorial Decision-Making