Finding mixed-strategy equilibria of continuous-action games without gradients using randomized policy networks.
Carlos MartinTuomas SandholmPublished in: CoRR (2022)
Keyphrases
- mixed strategy
- pure strategy
- nash equilibrium
- nash equilibria
- continuous action
- policy search
- finite horizon
- game theory
- incomplete information
- game theoretic
- solution concepts
- stochastic games
- continuous state
- optimal policy
- partially observable markov decision processes
- infinite horizon
- reinforcement learning
- pure nash equilibria
- control policies
- action space
- reward function
- action selection
- worst case