public void RecognitionNumberBasic()
{
int[] NeuronsInLayers = new int[] { number[0].Length, 10 };
HebbNetwork network = new HebbNetwork(NeuronsInLayers, 0.3);
for (int i = 0; i < number.Length; i++)
{
double[] target = Enumerable.Range(1, 10).Select(x => - 1d).ToArray();
target[i] = 1;
network.Train(number[i], target);
}
bool[] iscurrent = new bool[number.Length];
string wyniki = "";
for (int i = 0; i < number.Length; i++)
{
double[] res = network.Query(number[i]);
var max = res.Max();
var f = res.ToList().IndexOf(max);
iscurrent[i] = (f == i);
wyniki += "\n" + i + ":" + f;
//throw new Exception("Not Learned network. Errored in " + i);
}
if (((double)iscurrent.Count(x => x) / (double)iscurrent.Length) < 0.7)
{
throw new Exception("Ta siec jes tdo dupy wyniki:" + wyniki);
}
}
static void Main(string[] args)
{
//Graph.Data[] dates = new Graph.Data[]
//{
// new Graph.Data(new double[]{2,3,4,5,1,2},new double[]{0.1,2,4,5,10,11},"dwie tablice"),
// new Graph.Data(new double[]{1,3,4,6,8},"jedna tablica",GraphTypes.LINE),
// new Graph.Data(x =>2*x+1,-10,2,5 ,"funkcja liniowa"),
//};
int[] NeuronsInLayers = new int[] { number[0].Length, 10 };
HebbNetwork network = new HebbNetwork(NeuronsInLayers, 0.3);
for (int i = 0; i < number.Length; i++)
{
double[] target = Enumerable.Range(1, 10).Select(x => - 1d).ToArray();
target[i] = 1;
network.Train(number[i], target);
}
bool[] iscurrent = new bool[number.Length];
string wyniki = "";
for (int i = 0; i < number.Length; i++)
{
double[] res = network.Query(number[i]);
var max = res.Max();
var f = res.ToList().IndexOf(max);
iscurrent[i] = (f == i);
Console.WriteLine(f + "==" + i);
wyniki += "\n" + i + ":" + f;
//throw new Exception("Not Learned network. Errored in " + i);
}
if (((double)iscurrent.Count(x => x) / (double)iscurrent.Length) < 0.7)
{
throw new Exception("Ta siec jes tdo dupy wyniki:" + wyniki);
}
}