Off-Policy Primal-Dual Safe Reinforcement Learning.
Zifan WuBo TangQian LinChao YuShangqin MaoQianlong XieXingxing WangDong WangPublished in: CoRR (2024)
Keyphrases
- primal dual
- reinforcement learning
- linear programming
- affine scaling
- linear program
- convex optimization
- linear programming problems
- convergence rate
- interior point methods
- simplex algorithm
- approximation algorithms
- interior point algorithm
- semidefinite programming
- algorithm for linear programming
- learning algorithm
- interior point
- infeasible interior point
- machine learning
- duality gap
- markov decision processes
- optimal policy
- convex programming
- dynamic programming
- state space
- simplex method
- variational inequalities
- valid inequalities
- convex optimization problems
- dual formulation
- optimization problems
- convex functions
- step size
- np hard
- evolutionary algorithm
- image processing
- genetic algorithm