void CalcSuccess()
{
int[] inputs = new int[_network.InputLength];
int output;
double error = 0;
int expCount = 5000;
for (int i = 0; i < expCount; i++)
{
for (int j = 1; j < inputs.Length; j++)
{
inputs [j] = _rand.Next() % 100;
}
inputs[0] = 1;
output = _network.Think(inputs, null);
if (output != 0)
{
error += 1;
}
}
error /= expCount;
CompositionRoot.Instance.ExecuteInMainThread(() => {
_errorLabel.text = error.ToString("00.000");
});
}
private float GetError()
{
NeuralDecisionNetwork decider = _decider.Clone();
int successes = 0;
for (int i = 0; i < _trainingModels.Count; i++)
{
if (_trainingModels [i].Output == decider.Think(_trainingModels [i].Inputs, _trainingModels [i].Options))
{
successes++;
}
}
return((_trainingModels.Count - successes) / (float)_trainingModels.Count);
}
private int GetDecisionInd(DecisionType type, Game game, PlayerModel player, List <int> randoms, int points, Resource receivedRecource, WhereToGo whereToGo)
{
List <int> optionInds = AINeuralPlayer.GetOptionInds(type, game, _model, randoms, points, receivedRecource, whereToGo);
int[] inputs = AINeuralPlayer.GetInputs(type, game, _model, receivedRecource, whereToGo);
double[] inputsDouble = new double[inputs.Length];
for (int i = 0; i < inputs.Length; i++)
{
inputsDouble [i] = inputs [i];
}
NeuralDecisionNetwork chooser = GetChooserDecider(type);
int decisionInd = chooser.Think(inputs, optionInds);
return(decisionInd);
}
