private MDPUtilityFunction <TState> ValueDetermination(
IList <MDPTransition <TState, TAction> > validTransitions,
double gamma)
{
MDPUtilityFunction <TState> uf = utilityFunction.Copy();
double additional = 0.0;
if (validTransitions.Count > 0)
{
TState initState = validTransitions[0].GetInitialState();
double reward = MDP.GetRewardFor(initState);
additional =
validTransitions.Sum(
transition =>
MDP.GetTransitionProbability(transition) *
this.utilityFunction.GetUtility(transition.GetDestinationState()));
uf.SetUtility(initState, reward + (gamma * additional));
}
return(uf);
}
public MDPAgent(MDP <TState, TAction> mdp)
{
this.MDP = mdp;
this.CurrentState = mdp.GetInitialState();
this.CurrentReward = mdp.GetRewardFor(this.CurrentState);
}