DecPOMDP Usage

Decentralized POMDP

The DecPOMDP struct gives the following objects:

• γ: discount factor
• ℐ: agents
• 𝒮: state space
• 𝒜: joint action space
• 𝒪: joint observation space
• T: transition function
• O: joint observation function
• R: joint reward function

The agents ℐ are the players of the game. The joint action space 𝒜 is the set of all possible ordered pairs of actions amongst all of the agents. The joint observation space 𝒪 is the set of all possible joint observations. The transition function takes in a state s in 𝒮, a joint action a and a new state s'and returns the transition probability of going from s to s' by taking action a. The joint observation function takes in a state, s, a joint action, a, and a joint observation o in 𝒪 and returns a probability of observing o by taking action a from state s. The joint reward function R takes a state and a joint action in 𝒜 and returns a reward value.