public static GreedyGridWorldPolicy Create(
GridWorld world,
GridWorldValueTable gridWorldValueTable,
IGridWorldRewarder gridWorldRewarder)
{
var greedyPolicy = new GreedyGridWorldPolicy();
foreach (var state in world.AllStates())
{
greedyPolicy._actions[state] = FindBestAction(world, state, gridWorldValueTable, gridWorldRewarder);
}
return(greedyPolicy);
}
private static GridWorldAction FindBestAction(
GridWorld world,
GridWorldState state,
GridWorldValueTable gridWorldValueTable,
IGridWorldRewarder gridWorldRewarder)
{
var max = double.MinValue;
var maxAction = GridWorldAction.Down;
foreach (var action in world.AvailableActions(state))
{
var nextState = world.NextState(state, action);
var nextStateValue = gridWorldValueTable.Value(nextState);
var reward = gridWorldRewarder.Reward(state, action);
if (reward + nextStateValue > max)
{
max = reward + nextStateValue;
maxAction = action;
}
}
return(maxAction);
}
public GridWorldValueTable(GridWorld world)
{
_world = world;
_values = new double[world.AllStates().Count()];
}