public void Logistic()
{
double actual;
actual = new LogisticLoss(new[] { 0, 1, -0.5, 0.8 }).Loss(new[] { 0, 1, -0.5, 0.8 });
Assert.AreEqual(2.8938223431265548, actual, 1e-10);
actual = new LogisticLoss(new[] { 0, 1, -0.5, 0.8 }).Loss(new[] { 0, 1, 0.5, 0.8 });
Assert.AreEqual(3.2544961033487954, actual, 1e-10);
actual = new LogisticLoss(new[] { 0, 1, -0.5, 0.8 }).Loss(new[] { -5, 1, 0.5, 0.8 });
Assert.AreEqual(3.2544961033487954, actual, 1e-10);
actual = new LogisticLoss(new[] { 5.4, 1, -0.5, 0.8 }).Loss(new[] { -5.2, 1, 0.5, 0.8 });
Assert.AreEqual(42.765372851511813, actual, 1e-10);
actual = new LogisticLoss(new int[] { 0, 1, 0, 0 }).Loss(new double[] { 1, 1, 1, 1 });
Assert.AreEqual(3.250459373221156, actual, 1e-10);
actual = new LogisticLoss(new int[] { 0, 1, 0, 0 }).Loss(new double[] { 0, 0, 0, 0 });
Assert.AreEqual(4, actual);
actual = new LogisticLoss(new int[] { -1, 1, -1, -1 }).Loss(new double[] { -1, -1, -1, -1 });
Assert.AreEqual(6.1358494549990832, actual, 1e-10);
actual = new LogisticLoss(new int[] { 0, 0, 0, 1 }).Loss(new double[] { -1, 1, 1, 1 });
Assert.AreEqual(4.6931544141101194, actual, 1e-10);
actual = new LogisticLoss(new int[] { 0, 0, 0, 1 }).Loss(new double[] { -1, -1, -1, 1 });
Assert.AreEqual(7.5785444958880461, actual, 1e-10);
actual = new LogisticLoss(new int[] { -1, -1, -1, 1 }).Loss(new double[] { 0, 0, 0, 1 });
Assert.AreEqual(4.8946361239720115, actual, 1e-10);
actual = new LogisticLoss(new double[] { 0, 1, 0, 0 }).Loss(new double[] { 0, 0, 0, 0 });
Assert.AreEqual(4, actual);
Assert.AreEqual(0.31326168751822286d, new LogisticLoss().Loss(1, 1), 1e-10);
Assert.AreEqual(1.3132616875182228d, new LogisticLoss().Loss(-1, 1), 1e-10);
Assert.AreEqual(1.3132616875182228d, new LogisticLoss().Loss(1, -1), 1e-10);
Assert.AreEqual(0.31326168751822286d, new LogisticLoss().Loss(-1, -1), 1e-10);
Assert.AreEqual(0.0067153484891179669d, new LogisticLoss().Loss(1, 5), 1e-10);
Assert.AreEqual(6.0024756851377301d, new LogisticLoss().Loss(-1, 6), 1e-10);
Assert.AreEqual(7.0009114664537737d, new LogisticLoss().Loss(1, -7), 1e-10);
Assert.AreEqual(0.00033540637289566238d, new LogisticLoss().Loss(-1, -8), 1e-10);
Assert.AreEqual(1.3132616875182228d, new LogisticLoss().Loss(-1, 1), 1e-10);
Assert.AreEqual(0.69314718055994529d, new LogisticLoss().Loss(1, 0), 1e-10);
Assert.AreEqual(0.69314718055994529d, new LogisticLoss().Loss(-1, 0), 1e-10);
Assert.AreEqual(0.0067153484891179669d, new LogisticLoss().Loss(true, 5), 1e-10);
Assert.AreEqual(6.0024756851377301d, new LogisticLoss().Loss(false, 6), 1e-10);
Assert.AreEqual(7.0009114664537737d, new LogisticLoss().Loss(true, -7), 1e-10);
Assert.AreEqual(0.00033540637289566238d, new LogisticLoss().Loss(false, -8), 1e-10);
Assert.AreEqual(1.3132616875182228d, new LogisticLoss().Loss(false, 1), 1e-10);
Assert.AreEqual(0.69314718055994529d, new LogisticLoss().Loss(true, 0), 1e-10);
Assert.AreEqual(0.69314718055994529d, new LogisticLoss().Loss(false, 0), 1e-10);
}
public static void Run()
{
// Create
var trainX = new NDArray(new float[] { 0, 0, 0, 1, 1, 0, 1, 1 }).Reshape(4, 2);
var trainY = new NDArray(new float[] { 0, 1, 1, 0 });
var batch_size = 2;
var train_data = new NDArrayIter(trainX, trainY, batch_size);
var val_data = new NDArrayIter(trainX, trainY, batch_size);
var net = new Sequential();
net.Add(new Dense(64, ActivationType.Relu));
net.Add(new Dense(1));
var gpus = TestUtils.ListGpus();
var ctxList = gpus.Count > 0 ? gpus.Select(x => Context.Gpu(x)).ToArray() : new[] { Context.Cpu() };
net.Initialize(new Uniform(), ctxList.ToArray());
var trainer = new Trainer(net.CollectParams(), new Adam());
var epoch = 1000;
var metric = new BinaryAccuracy();
var binary_crossentropy = new LogisticLoss();
float lossVal = 0;
for (var iter = 0; iter < epoch; iter++)
{
train_data.Reset();
lossVal = 0;
while (!train_data.End())
{
var batch = train_data.Next();
var data = Utils.SplitAndLoad(batch.Data[0], ctxList);
var label = Utils.SplitAndLoad(batch.Label[0], ctxList);
NDArrayList outputs = null;
using (var ag = Autograd.Record())
{
outputs = Enumerable.Zip(data, label, (x, y) =>
{
var z = net.Call(x);
NDArray loss = binary_crossentropy.Call(z, y);
loss.Backward();
lossVal += loss.Mean();
return(z);
}).ToList();
}
metric.Update(label, outputs.ToArray());
trainer.Step(batch.Data[0].Shape[0]);
}
var(name, acc) = metric.Get();
metric.Reset();
Console.WriteLine($"Loss: {lossVal}");
Console.WriteLine($"Training acc at epoch {iter}: {name}={acc * 100}%");
}
}
