private static GamblersWorldAction FindBestAction(
GamblersWorld world,
GamblersWorldState state,
GamblersValueTable valueTable,
IGamblersWorldRewarder rewarder)
{
var maxActionValue = double.MinValue;
var maxAction = new GamblersWorldAction(0);
foreach (var action in world.AvailableActions(state))
{
var actionValue = 0.0;
foreach (var(nextState, pNextState) in world.PossibleStates(state, action))
{
var nextStateValue = valueTable.Value(nextState);
var reward = rewarder.Reward(state, nextState, action);
actionValue += pNextState * (reward + nextStateValue);
}
if (actionValue > maxActionValue)
{
maxActionValue = actionValue;
maxAction = action;
}
}
return(maxAction);
}
public static GreedyGamblersPolicy Create(
GamblersWorld world,
GamblersValueTable valueTable,
IGamblersWorldRewarder rewarder)
{
var greedyPolicy = new GreedyGamblersPolicy(world);
foreach (var state in world.AllStates())
{
var bestAction = FindBestAction(world, state, valueTable, rewarder);
greedyPolicy._actions[state.DollarsInHand] = bestAction.Stake;
}
return(greedyPolicy);
}