A Unified Worst Case for Classical Simplex and Policy Iteration Pivot Rules.
Yann DisserNils MosisPublished in: ISAAC (2023)
Keyphrases
- policy iteration
- worst case
- markov decision processes
- simplex method
- model free
- sample path
- linear programming
- reinforcement learning
- fixed point
- least squares
- lower bound
- optimal policy
- upper bound
- markov decision process
- infinite horizon
- convergence rate
- finite state
- policy evaluation
- np hard
- temporal difference
- optimal control
- average reward
- multi agent
- data mining
- markov decision problems
- optical flow