public void TanhNodeCanFigureOutOr()
{
var description = new NetDescription
{
Nodes = new[] {
new NodeDescription
{
NodeId = 0, Weight = 0.001f,
Aggregator = "sum", Processor = "tanh",
Inputs = new []
{
new NodeInputDescription
{
FromInputVector = true, InputId = 0, Weight = -0.001f
},
new NodeInputDescription
{
FromInputVector = true, InputId = 1, Weight = 0.001f
}
}
}
},
Outputs = new[] { 0 }
};
var net = Net.FromDescription(description);
var tests = new[]
{
Tuple.Create(new[] { 1f, 0f }, new[] { 1f }),
Tuple.Create(new[] { 1f, 1f }, new[] { 1f }),
Tuple.Create(new[] { 0f, 1f }, new[] { 1f }),
Tuple.Create(new[] { 0f, 0f }, new[] { -1f })
};
var trainer = new SimpleTrainer();
var loss = trainer.Train(
net: net,
tests: tests,
desiredError: 0.01f,
maxEpochs: 50000,
learningRate: 0.5f);
Assert.IsTrue(loss < 0.01f);
}
public void TurtleZebraGiraffe()
{
var tests = new[]
{
Tuple.Create(new[] { 1f, 1f, 1f, 1f, 1f, 1f, 1f, 1f, 1f, 1f }, new[] { 0f, 0f, 1f }),
Tuple.Create(new[] { 1f, -1f, -1f, -1f, -1f, -1f, -1f, -1f, -1f, -1f }, new[] { 1f, 0f, 0f }),
Tuple.Create(new[] { 1f, 1f, 1f, 1f, -1f, -1f, -1f, -1f, -1f, -1f }, new[] { 0f, 1f, 0f })
};
var description = SimpleDescriptionBuilder.GetDescription(10, new[] { 5, 5, 5, 3 });
foreach (var node in description.Nodes.Where(node => description.Outputs.Contains(node.NodeId)))
{
node.Processor = "sigmoid";
}
var net = Net.FromDescription(description);
WeightFiller.FillWeights(net, .005f);
var trainer = new SimpleTrainer();
var error = trainer.Train(
net: net,
tests: tests,
desiredError: 0.01f,
maxEpochs: 20000,
learningRate: .5f);
Assert.IsTrue(error < .01f);
var eval = net.GetEvaluationFunction();
foreach (var test in tests)
{
var output = eval(test.Item1);
Console.WriteLine(string.Join(",", output));
}
}
public void TanhNetCanBeTrainedOnXOr()
{
var tests = new[]
{
Tuple.Create(new[] { 1f, -1f }, new[] { 1f }),
Tuple.Create(new[] { 1f, 1f }, new[] { -1f }),
Tuple.Create(new[] { -1f, 1f }, new[] { 1f }),
Tuple.Create(new[] { -1f, -1f }, new[] { -1f })
};
var initialDescription = new NetDescription
{
Nodes = new[]
{
new NodeDescription
{
NodeId = 0, Aggregator = "sum", Processor = "tanh",
Weight = .21f, Inputs = new []
{
new NodeInputDescription
{
FromInputVector = true, InputId = 0, Weight = -.07f
},
new NodeInputDescription
{
FromInputVector = true, InputId = 1, Weight = -.28f
}
}
},
new NodeDescription
{
NodeId = 1, Aggregator = "sum", Processor = "tanh",
Weight = -.29f, Inputs = new []
{
new NodeInputDescription
{
FromInputVector = true, InputId = 0, Weight = .41f
},
new NodeInputDescription
{
FromInputVector = true, InputId = 1, Weight = -.05f
}
}
},
new NodeDescription
{
NodeId = 2, Aggregator = "sum", Processor = "tanh",
Weight = .11f, Inputs = new []
{
new NodeInputDescription
{
FromInputVector = false, InputId = 0, Weight = -.1f
},
new NodeInputDescription
{
FromInputVector = false, InputId = 1, Weight = -.21f
}
}
}
},
Outputs = new[] { 2 }
};
var net = Net.FromDescription(initialDescription);
WeightFiller.FillWeights(net, .05f);
var trainer = new SimpleTrainer();
var error = trainer.Train(
net: net,
tests: tests,
desiredError: .001f,
maxEpochs: 100000,
learningRate: 5f);
Console.WriteLine(error);
Assert.IsTrue(error < 0.1f);
}
static void Main(string[] args)
{
var events = ReadEventFile("trainingEvents.csv");
var ins = events
.Where(e => (e.Open - e.NextLow) / e.Open > .1f)
.Select(evts => evts.GetInputArray());
Console.WriteLine($"Qualified Events: {ins.Count()}");
var unsupervisedTests = ins.Select(i => Tuple.Create(i, i));
var supervisedTests = events.Select(evt => Tuple.Create(evt.GetInputArray(), evt.GetOutputArray()));
var builder = new LayerBuilder();
var description = builder.BuildDescription(5, new[]
{
new LayerBuilder.LayerSpec(10, "sum", "softplus"),
new LayerBuilder.LayerSpec(10, "sum", "softplus"),
new LayerBuilder.LayerSpec(10, "sum", "tanh"),
new LayerBuilder.LayerSpec(10, "sum", "softplus"),
new LayerBuilder.LayerSpec(10, "sum", "softplus"),
new LayerBuilder.LayerSpec(5, "sum", null)
});
//var description = builder.BuildDescription(5, new[]
//{
// new LayerBuilder.LayerSpec(5, "sum", "softplus"),
// new LayerBuilder.LayerSpec(6, "sum", "tanh"),
// new LayerBuilder.LayerSpec(4, "sum", "softplus"),
// new LayerBuilder.LayerSpec(5, "sum", null)
//});
var net = Net.FromDescription(description);
var trainer = new SimpleTrainer();
trainer.Train(
net: net,
tests: unsupervisedTests,
desiredError: 5e-6f,
maxEpochs: 200,
learningRate: 0.75f);
trainer.Train(
net: net,
tests: unsupervisedTests,
desiredError: 5e-6f,
maxEpochs: 9500,
learningRate: 0.5f);
trainer.Train(
net: net,
tests: unsupervisedTests,
desiredError: 1e-8f,
maxEpochs: 200,
learningRate: .25f);
trainer.Train(
net: net,
tests: unsupervisedTests,
desiredError: 1e-8f,
maxEpochs: 200,
learningRate: .125f);
trainer.Train(
net: net,
tests: unsupervisedTests,
desiredError: 1e-8f,
maxEpochs: 200,
learningRate: .0625f);
for (var i = 0; i < 5; i++)
{
var nextDescription = net.Description;
var firstSigmoidId = nextDescription.Nodes.First(n => n.Processor == "tanh").NodeId;
nextDescription.Nodes = nextDescription.Nodes.Where(n => n.NodeId != firstSigmoidId).ToArray();
foreach (var node in nextDescription.Nodes)
{
node.Inputs = node.Inputs
.Where(inp => inp.FromInputVector || inp.InputId != firstSigmoidId)
.ToArray();
}
net = Net.FromDescription(nextDescription);
Console.WriteLine($"Removed {i + 1} sigmoids");
trainer.Train(
net: net,
tests: unsupervisedTests,
desiredError: 1e-6f,
maxEpochs: 400,
learningRate: 0.5f);
}
trainer.Train(
net: net,
tests: unsupervisedTests,
desiredError: 1e-6f,
maxEpochs: 500,
learningRate: 0.5f);
trainer.Train(
net: net,
tests: unsupervisedTests,
desiredError: 1e-6f,
maxEpochs: 400,
learningRate: 0.25f);
var finalDescription = net.Description;
var outsRemoved = new[] { finalDescription.Outputs[1], finalDescription.Outputs[2], finalDescription.Outputs[3], finalDescription.Outputs[4] };
finalDescription.Nodes = finalDescription.Nodes
.Where(n => !outsRemoved.Contains(n.NodeId))
.ToArray();
finalDescription.Outputs = new[] { finalDescription.Outputs[0] };
var nextNet = Net.FromDescription(finalDescription);
trainer.Train(
net: nextNet,
tests: supervisedTests,
desiredError: 1e-8f,
maxEpochs: 20000,
learningRate: .25f);
trainer.Train(
net: nextNet,
tests: supervisedTests,
desiredError: 1e-8f,
maxEpochs: 20000,
learningRate: .125f);
trainer.Train(
net: nextNet,
tests: supervisedTests,
desiredError: 1e-8f,
maxEpochs: 2000,
learningRate: .0625f);
trainer.Train(
net: nextNet,
tests: supervisedTests,
desiredError: 1e-8f,
maxEpochs: 2000,
learningRate: .03125f);
var final = net.Description;
var finalText = JsonConvert.SerializeObject(final);
using (var writer = File.CreateText("out3.json"))
{
writer.Write(finalText);
}
Console.WriteLine();
Console.ReadLine();
}
