public void Mfn_LearningTestMethod()
{
double[,] w = new double[2, 3];
w[0, 0] = 0.1;
w[0, 1] = -0.2;
w[0, 2] = 0.1;
w[1, 0] = 0.1;
w[1, 1] = -0.1;
w[1, 2] = 0.3;
Mfn network = new Mfn(w, Activation.Sigmoid, ActivationD.Sigmoid);
double[,] w1 = new double[1, 3];
w1[0, 0] = 0.2;
w1[0, 1] = 0.2;
w1[0, 2] = 0.3;
network.AddLayer(w1, Activation.Sigmoid, ActivationD.Sigmoid);
PairValue test = new PairValue();
test.Input = new double[2];
test.Input[0] = 0.9;
test.Input[0] = 0.1;
test.Output = new double[1];
test.Output[0] = 0.9;
double actual = network.Learning(new PairValue[1] {
test
});
double expected = 0.039720;
Assert.AreEqual(expected, actual, 0.000001);
}
public void SaveLoadBinMfnTestMethod()
{
double[,] w = new double[2, 3];
w[0, 0] = 1.5;
w[0, 1] = -1.0;
w[0, 2] = -1.0;
w[1, 0] = 0.5;
w[1, 1] = -1.0;
w[1, 2] = -1.0;
Mfn network = new Mfn(w, Activation.Threshold, 0);
double[,] w1 = new double[1, 3];
w1[0, 0] = -0.5;
w1[0, 1] = 1.0;
w1[0, 2] = -1.0;
network.AddLayer(w1, Activation.Threshold, 0);
network.Save(@"C:\Users\super_be\Source\Repos\Neuron.Net\Neuron.Net\Neuron.Net\Neuron.Net.Core.Tests\Data\TestNN.nnn");
Mfn nn = Mfn.Load(@"C:\Users\super_be\Source\Repos\Neuron.Net\Neuron.Net\Neuron.Net\Neuron.Net.Core.Tests\Data\TestNN.nnn");
double[] x = new double[2];
x[0] = 1;
x[1] = 1;
double[] result = nn.Calc(x);
double actual = result[0];
double expected = 0;
Assert.AreEqual(expected, actual);
}
public void Mfn_2_2_1_TestMethod()
{
double[,] w = new double[2, 3];
w[0, 0] = 1.5;
w[0, 1] = -1.0;
w[0, 2] = -1.0;
w[1, 0] = 0.5;
w[1, 1] = -1.0;
w[1, 2] = -1.0;
Mfn network = new Mfn(w, Activation.Threshold, 0);
double[,] w1 = new double[1, 3];
w1[0, 0] = -0.5;
w1[0, 1] = 1.0;
w1[0, 2] = -1.0;
network.AddLayer(w1, Activation.Threshold, 0);
double[] x = new double[2];
x[0] = 1;
x[1] = 1;
double[] result = network.Calc(x);
double actual = result[0];
double expected = 0;
Assert.AreEqual(expected, actual);
x[0] = 1;
x[1] = 0;
result = network.Calc(x);
actual = result[0];
expected = 1;
Assert.AreEqual(expected, actual);
x[0] = 0;
x[1] = 1;
result = network.Calc(x);
actual = result[0];
expected = 1;
Assert.AreEqual(expected, actual);
x[0] = 0;
x[1] = 0;
result = network.Calc(x);
actual = result[0];
expected = 0;
Assert.AreEqual(expected, actual);
}
public void Mfn_Learning_2_2_2_1_100_TestMethod()
{
double[,] w = new double[2, 3];
w[0, 0] = 2;
w[0, 1] = -2;
w[0, 2] = -2;
w[1, 0] = 2;
w[1, 1] = 3;
w[1, 2] = 3;
Mfn network = new Mfn(w, Activation.Sigmoid, ActivationD.Sigmoid);
double[,] w1 = new double[2, 3];
w1[0, 0] = 3;
w1[0, 1] = -2;
w1[0, 2] = -4;
w1[1, 0] = -2;
w1[1, 1] = 2;
w1[1, 2] = 2;
network.AddLayer(w1, Activation.Sigmoid, ActivationD.Sigmoid);
double[,] w2 = new double[1, 3];
w2[0, 0] = -2;
w2[0, 1] = 3;
w2[0, 2] = 1;
network.AddLayer(w2, Activation.Sigmoid, ActivationD.Sigmoid);
PairValue test = new PairValue();
test.Input = new double[2];
test.Input[0] = 0.1;
test.Input[0] = 0.9;
test.Output = new double[1];
test.Output[0] = 0.9;
Mfn nn = network.Learning(1000, new PairValue[1] {
test
}, 0.8, 0, 0.000001);
double[] result = nn.Calc(test.Input);
double actual = result[0];
double expected = 0.899924294265539;
Assert.AreEqual(expected, actual, 0.000001);
}
public void Mfn_Learning_2_2_2_1_TestMethod()
{
double[,] w = new double[2, 3];
w[0, 0] = 2;
w[0, 1] = -2;
w[0, 2] = -2;
w[1, 0] = 2;
w[1, 1] = 3;
w[1, 2] = 3;
Mfn network = new Mfn(w, Activation.Sigmoid, ActivationD.Sigmoid);
double[,] w1 = new double[2, 3];
w1[0, 0] = 3;
w1[0, 1] = -2;
w1[0, 2] = -4;
w1[1, 0] = -2;
w1[1, 1] = 2;
w1[1, 2] = 2;
network.AddLayer(w1, Activation.Sigmoid, ActivationD.Sigmoid);
double[,] w2 = new double[1, 3];
w2[0, 0] = -2;
w2[0, 1] = 3;
w2[0, 2] = 1;
network.AddLayer(w2, Activation.Sigmoid, ActivationD.Sigmoid);
PairValue test = new PairValue();
test.Input = new double[2];
test.Input[0] = 0.1;
test.Input[0] = 0.9;
test.Output = new double[1];
test.Output[0] = 0.9;
double actual = network.Learning(new PairValue[1] {
test
}, 0.8, 0);
double expected = 0.0289141550069473;
Assert.AreEqual(expected, actual, 0.000001);
}
public string Encode
(
string delimiter = IrbisText.IrbisDelimiter
)
{
StringBuilder result = new StringBuilder(512);
result.Append(Mfn.ToInvariantString())
.Append('#')
.Append(((int)Status).ToInvariantString())
.Append(delimiter)
.Append("0#")
.Append(Version.ToInvariantString())
.Append(delimiter);
foreach (var field in Fields)
{
result.Append(field)
.Append(delimiter);
}
return(result.ToString());
} // method Encode
public void Mfn_Number_TestMethod()
{
Perceptron p63 = new Perceptron(63, 6, Activation.Threshold, 0, ActivationD.Threshold);
Mfn network = new Mfn(p63);
Perceptron p6 = new Perceptron(6, 9, Activation.Threshold, 0, ActivationD.Threshold);
network.AddLayer(p6);
//Perceptron p63 = new Perceptron(63, 9, Activation.Threshold, 0.5, ActivationD.Threshold);
//Mfn network = new Mfn(p63);
PairValue t0 = new PairValue();
t0.Input = new double[63] {
0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 1.0, 1.0, 1.0, 1.0, 1.0, 0.0, 0.0, 1.0, 0.0, 0.0, 0.0, 1.0, 0.0, 0.0, 1.0, 0.0, 0.0, 0.0, 1.0, 0.0, 0.0, 1.0, 0.0, 0.0, 0.0, 1.0, 0.0, 0.0, 1.0, 0.0, 0.0, 0.0, 1.0, 0.0, 0.0, 1.0, 0.0, 0.0, 0.0, 1.0, 0.0, 0.0, 1.0, 1.0, 1.0, 1.0, 1.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0
};
t0.Output = new double[9] {
0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0
};
PairValue t1 = new PairValue();
t1.Input = new double[63] {
0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 1.0, 0.0, 0.0, 0.0, 0.0, 0.0, 1.0, 1.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 1.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 1.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 1.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 1.0, 0.0, 0.0, 0.0, 0.0, 1.0, 1.0, 1.0, 1.0, 1.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0
};
t1.Output = new double[9] {
1.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0
};
PairValue t2 = new PairValue();
t2.Input = new double[63] {
0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 1.0, 1.0, 1.0, 1.0, 1.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 1.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 1.0, 0.0, 0.0, 1.0, 1.0, 1.0, 1.0, 1.0, 0.0, 0.0, 1.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 1.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 1.0, 1.0, 1.0, 1.0, 1.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0
};
t2.Output = new double[9] {
0.0, 1.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0
};
PairValue t3 = new PairValue();
t3.Input = new double[63] {
0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 1.0, 1.0, 1.0, 1.0, 1.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 1.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 1.0, 0.0, 0.0, 0.0, 0.0, 1.0, 1.0, 1.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 1.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 1.0, 0.0, 0.0, 1.0, 1.0, 1.0, 1.0, 1.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0
};
t3.Output = new double[9] {
0.0, 0.0, 1.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0
};
PairValue t4 = new PairValue();
t4.Input = new double[63] {
0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 1.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 1.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 1.0, 0.0, 1.0, 0.0, 0.0, 0.0, 0.0, 1.0, 1.0, 1.0, 1.0, 0.0, 0.0, 0.0, 0.0, 0.0, 1.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 1.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 1.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0
};
t4.Output = new double[9] {
0.0, 0.0, 0.0, 1.0, 0.0, 0.0, 0.0, 0.0, 0.0
};
PairValue t5 = new PairValue();
t5.Input = new double[63] {
0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 1.0, 1.0, 1.0, 1.0, 1.0, 0.0, 0.0, 1.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 1.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 1.0, 1.0, 1.0, 1.0, 1.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 1.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 1.0, 0.0, 0.0, 1.0, 1.0, 1.0, 1.0, 1.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0
};
t5.Output = new double[9] {
0.0, 0.0, 0.0, 0.0, 1.0, 0.0, 0.0, 0.0, 0.0
};
PairValue t6 = new PairValue();
t6.Input = new double[63] {
0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 1.0, 1.0, 1.0, 1.0, 1.0, 0.0, 0.0, 1.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 1.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 1.0, 1.0, 1.0, 1.0, 1.0, 0.0, 0.0, 1.0, 0.0, 0.0, 0.0, 1.0, 0.0, 0.0, 1.0, 0.0, 0.0, 0.0, 1.0, 0.0, 0.0, 1.0, 1.0, 1.0, 1.0, 1.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0
};
t6.Output = new double[9] {
0.0, 0.0, 0.0, 0.0, 0.0, 1.0, 0.0, 0.0, 0.0
};
PairValue t7 = new PairValue();
t7.Input = new double[63] {
0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 1.0, 1.0, 1.0, 1.0, 1.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 1.0, 0.0, 0.0, 0.0, 0.0, 0.0, 1.0, 0.0, 0.0, 0.0, 0.0, 0.0, 1.0, 0.0, 0.0, 0.0, 0.0, 0.0, 1.0, 0.0, 0.0, 0.0, 0.0, 0.0, 1.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 1.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0
};
t7.Output = new double[9] {
0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 1.0, 0.0, 0.0
};
PairValue t8 = new PairValue();
t8.Input = new double[63] {
0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 1.0, 1.0, 1.0, 1.0, 1.0, 0.0, 0.0, 1.0, 0.0, 0.0, 0.0, 1.0, 0.0, 0.0, 1.0, 0.0, 0.0, 0.0, 1.0, 0.0, 0.0, 1.0, 1.0, 1.0, 1.0, 1.0, 0.0, 0.0, 1.0, 0.0, 0.0, 0.0, 1.0, 0.0, 0.0, 1.0, 0.0, 0.0, 0.0, 1.0, 0.0, 0.0, 1.0, 1.0, 1.0, 1.0, 1.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0
};
t8.Output = new double[9] {
0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 1.0, 0.0
};
PairValue t9 = new PairValue();
t9.Input = new double[63] {
0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 1.0, 1.0, 1.0, 1.0, 1.0, 0.0, 0.0, 1.0, 0.0, 0.0, 0.0, 1.0, 0.0, 0.0, 1.0, 0.0, 0.0, 0.0, 1.0, 0.0, 0.0, 1.0, 1.0, 1.0, 1.0, 1.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 1.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 1.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 1.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0
};
t9.Output = new double[9] {
0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 1.0
};
//Mfn nn = network.Learning(1000, new PairValue[10] { t0, t1, t2, t3, t4, t5, t6, t7, t8, t9 });
Mfn nn = network.Learning(1000, new PairValue[10] {
t0, t1, t2, t3, t4, t5, t6, t7, t8, t9
}, 0.3, 0.7, 0.00001);
double[] result = nn.Calc(t0.Input);
double actual = result[0];
double expected = 0;
Assert.AreEqual(expected, actual);
actual = result[1];
expected = 0;
Assert.AreEqual(expected, actual);
actual = result[2];
expected = 0;
Assert.AreEqual(expected, actual);
actual = result[3];
expected = 0;
Assert.AreEqual(expected, actual);
actual = result[4];
expected = 0;
Assert.AreEqual(expected, actual);
actual = result[5];
expected = 0;
Assert.AreEqual(expected, actual);
actual = result[6];
expected = 0;
Assert.AreEqual(expected, actual);
actual = result[7];
expected = 0;
Assert.AreEqual(expected, actual);
actual = result[8];
expected = 0;
Assert.AreEqual(expected, actual);
result = nn.Calc(t1.Input);
actual = result[0];
expected = 1;
Assert.AreEqual(expected, actual);
actual = result[1];
expected = 0;
Assert.AreEqual(expected, actual);
actual = result[2];
expected = 0;
Assert.AreEqual(expected, actual);
actual = result[3];
expected = 0;
Assert.AreEqual(expected, actual);
actual = result[4];
expected = 0;
Assert.AreEqual(expected, actual);
actual = result[5];
expected = 0;
Assert.AreEqual(expected, actual);
actual = result[6];
expected = 0;
Assert.AreEqual(expected, actual);
actual = result[7];
expected = 0;
Assert.AreEqual(expected, actual);
actual = result[8];
expected = 0;
Assert.AreEqual(expected, actual);
result = nn.Calc(t2.Input);
actual = result[0];
expected = 0;
Assert.AreEqual(expected, actual);
actual = result[1];
expected = 1;
Assert.AreEqual(expected, actual);
actual = result[2];
expected = 0;
Assert.AreEqual(expected, actual);
actual = result[3];
expected = 0;
Assert.AreEqual(expected, actual);
actual = result[4];
expected = 0;
Assert.AreEqual(expected, actual);
actual = result[5];
expected = 0;
Assert.AreEqual(expected, actual);
actual = result[6];
expected = 0;
Assert.AreEqual(expected, actual);
actual = result[7];
expected = 0;
Assert.AreEqual(expected, actual);
actual = result[8];
expected = 0;
Assert.AreEqual(expected, actual);
result = nn.Calc(t3.Input);
actual = result[0];
expected = 0;
Assert.AreEqual(expected, actual);
actual = result[1];
expected = 0;
Assert.AreEqual(expected, actual);
actual = result[2];
expected = 1;
Assert.AreEqual(expected, actual);
actual = result[3];
expected = 0;
Assert.AreEqual(expected, actual);
actual = result[4];
expected = 0;
Assert.AreEqual(expected, actual);
actual = result[5];
expected = 0;
Assert.AreEqual(expected, actual);
actual = result[6];
expected = 0;
Assert.AreEqual(expected, actual);
actual = result[7];
expected = 0;
Assert.AreEqual(expected, actual);
actual = result[8];
expected = 0;
Assert.AreEqual(expected, actual);
result = nn.Calc(t4.Input);
actual = result[0];
expected = 0;
Assert.AreEqual(expected, actual);
actual = result[1];
expected = 0;
Assert.AreEqual(expected, actual);
actual = result[2];
expected = 0;
Assert.AreEqual(expected, actual);
actual = result[3];
expected = 1;
Assert.AreEqual(expected, actual);
actual = result[4];
expected = 0;
Assert.AreEqual(expected, actual);
actual = result[5];
expected = 0;
Assert.AreEqual(expected, actual);
actual = result[6];
expected = 0;
Assert.AreEqual(expected, actual);
actual = result[7];
expected = 0;
Assert.AreEqual(expected, actual);
actual = result[8];
expected = 0;
Assert.AreEqual(expected, actual);
result = nn.Calc(t5.Input);
actual = result[0];
expected = 0;
Assert.AreEqual(expected, actual);
actual = result[1];
expected = 0;
Assert.AreEqual(expected, actual);
actual = result[2];
expected = 0;
Assert.AreEqual(expected, actual);
actual = result[3];
expected = 0;
Assert.AreEqual(expected, actual);
actual = result[4];
expected = 1;
Assert.AreEqual(expected, actual);
actual = result[5];
expected = 0;
Assert.AreEqual(expected, actual);
actual = result[6];
expected = 0;
Assert.AreEqual(expected, actual);
actual = result[7];
expected = 0;
Assert.AreEqual(expected, actual);
actual = result[8];
expected = 0;
Assert.AreEqual(expected, actual);
result = nn.Calc(t6.Input);
actual = result[0];
expected = 0;
Assert.AreEqual(expected, actual);
actual = result[1];
expected = 0;
Assert.AreEqual(expected, actual);
actual = result[2];
expected = 0;
Assert.AreEqual(expected, actual);
actual = result[3];
expected = 0;
Assert.AreEqual(expected, actual);
actual = result[4];
expected = 0;
Assert.AreEqual(expected, actual);
actual = result[5];
expected = 1;
Assert.AreEqual(expected, actual);
actual = result[6];
expected = 0;
Assert.AreEqual(expected, actual);
actual = result[7];
expected = 0;
Assert.AreEqual(expected, actual);
actual = result[8];
expected = 0;
Assert.AreEqual(expected, actual);
result = nn.Calc(t7.Input);
actual = result[0];
expected = 0;
Assert.AreEqual(expected, actual);
actual = result[1];
expected = 0;
Assert.AreEqual(expected, actual);
actual = result[2];
expected = 0;
Assert.AreEqual(expected, actual);
actual = result[3];
expected = 0;
Assert.AreEqual(expected, actual);
actual = result[4];
expected = 0;
Assert.AreEqual(expected, actual);
actual = result[5];
expected = 0;
Assert.AreEqual(expected, actual);
actual = result[6];
expected = 1;
Assert.AreEqual(expected, actual);
actual = result[7];
expected = 0;
Assert.AreEqual(expected, actual);
actual = result[8];
expected = 0;
Assert.AreEqual(expected, actual);
result = nn.Calc(t8.Input);
actual = result[0];
expected = 0;
Assert.AreEqual(expected, actual);
actual = result[1];
expected = 0;
Assert.AreEqual(expected, actual);
actual = result[2];
expected = 0;
Assert.AreEqual(expected, actual);
actual = result[3];
expected = 0;
Assert.AreEqual(expected, actual);
actual = result[4];
expected = 0;
Assert.AreEqual(expected, actual);
actual = result[5];
expected = 0;
Assert.AreEqual(expected, actual);
actual = result[6];
expected = 0;
Assert.AreEqual(expected, actual);
actual = result[7];
expected = 1;
Assert.AreEqual(expected, actual);
actual = result[8];
expected = 0;
Assert.AreEqual(expected, actual);
result = nn.Calc(t9.Input);
actual = result[0];
expected = 0;
Assert.AreEqual(expected, actual);
actual = result[1];
expected = 0;
Assert.AreEqual(expected, actual);
actual = result[2];
expected = 0;
Assert.AreEqual(expected, actual);
actual = result[3];
expected = 0;
Assert.AreEqual(expected, actual);
actual = result[4];
expected = 0;
Assert.AreEqual(expected, actual);
actual = result[5];
expected = 0;
Assert.AreEqual(expected, actual);
actual = result[6];
expected = 0;
Assert.AreEqual(expected, actual);
actual = result[7];
expected = 0;
Assert.AreEqual(expected, actual);
actual = result[8];
expected = 1;
Assert.AreEqual(expected, actual);
}
public void Mfn_2_2_3_TestMethod()
{
double[,] w = new double[2, 3];
w[0, 0] = 1.0;
w[0, 1] = 2.0;
w[0, 2] = -3.0;
w[1, 0] = -2.0;
w[1, 1] = 0.5;
w[1, 2] = 1.0;
Mfn network = new Mfn(w, Activation.Linear);
double[,] w1 = new double[3, 3];
w1[0, 0] = 2.0;
w1[0, 1] = -1.0;
w1[0, 2] = -3.0;
w1[1, 0] = 1.0;
w1[1, 1] = 5.0;
w1[1, 2] = 1.0;
w1[2, 0] = 3.0;
w1[2, 1] = 4.0;
w1[2, 2] = 2.0;
network.AddLayer(w1, Activation.Linear);
double[] x = new double[2];
x[0] = 0.0;
x[1] = 1.0;
double[] result = network.Calc(x);
double actual = result[0];
double expected = 7.0;
Assert.AreEqual(expected, actual);
actual = result[1];
expected = -10.0;
Assert.AreEqual(expected, actual);
actual = result[2];
expected = -7.0;
Assert.AreEqual(expected, actual);
double[,] w2 = new double[3, 3];
w2[0, 0] = 5.0;
w2[0, 1] = -3.5;
w2[0, 2] = 2.0;
w2[1, 0] = 3.0;
w2[1, 1] = 10.5;
w2[1, 2] = -13.0;
w2[2, 0] = 3.0;
w2[2, 1] = 9.0;
w2[2, 2] = -10.0;
Mfn network1 = new Mfn(w2, Activation.Linear);
x = new double[2];
x[0] = 0.0;
x[1] = 1.0;
result = network1.Calc(x);
actual = result[0];
expected = 7.0;
Assert.AreEqual(expected, actual);
actual = result[1];
expected = -10.0;
Assert.AreEqual(expected, actual);
actual = result[2];
expected = -7.0;
Assert.AreEqual(expected, actual);
}
