private MDPState descendByUCB(MDPState state)
{
RepairActionsSet actions = repairActionSearcher.ComputePossibleAcions(state.State);
if (actions == null)
{
return(null);
}
RepairAction currentAction = actions.NextAction();
double U;
RepairAction bestAction = null;
double ans = double.MinValue;
while (currentAction != null && currentAction.Count != 0)
{
int actionNb = state.GetNumOfTimesChosen(currentAction);
if (actionNb == 0)
{
U = (-costEstimator.FPCost(currentAction, state.State.HealthState)) / (currentAction.Count);
}
else
{
double alpha = state.GetStateActionValue(currentAction);
U = alpha + numOfDiag * (Math.Sqrt(Math.Log10(state.NumberOfVisits) / actionNb));
}
if (U > ans) //update best action
{
bestAction = currentAction;
ans = U;
}
currentAction = actions.NextAction();
}
MDPState nextState = null;
if (bestAction != null)
{
state.LatestChosenAction = bestAction;
nextState = ComputeNextState(state, bestAction);
}
state.NumberOfVisits++;
return(nextState);
}
public override RepairAction Plan(SystemState state)
{
ResetProperties();
if (state == null || state.Diagnoses == null || state.Diagnoses.Count == 0)
{
return(null);
}
RepairActionsSet actions = repairActionSearcher.ComputePossibleAcions(state);
if (actions == null)
{
return(null);
}
double min = double.MaxValue;
RepairAction bestRepairAction = null;
RepairAction action = actions.NextAction();
int counter = 0;
while (action != null && action.Count > 0)
{
counter++;
if (Stop())
{
break;
}
double val = costEstimator.WastedCostUtility(action, state);
if (val < min)
{
min = val;
bestRepairAction = action;
}
action = actions.NextAction();
}
FillIterationDetails(bestRepairAction, counter, min);
return(bestRepairAction);
}
