Esempio n. 1
0
    private void Evaluate(float[] actions)
    {
        float maxValue = actions.Max();
        int   maxIndex = actions.ToList().IndexOf(maxValue);

        float[] action = { maxIndex };
        var     eval   = _ball.Action(action);

        _currentReward += eval.Item1;
        if (eval.Item2)
        {
            Debug.Log("Accumulated Reward: " + _currentReward);
            _currentReward = 0;
            _ball.Reset();
        }
    }
Esempio n. 2
0
 public override void AgentReset()
 {
     _nvbManager.Reset();
 }