如何测试用PyBrain开发的神经网络

Question

I'm new to machine learning and am trying to learn how to develop neural networks for prediction purposes in Python. I followed a basic tutorial from PyBrain and have successfully set up a neural network and have trained it (supervised learning). Here is the code:

ds = SupervisedDataSet(2, 1)
ds.addSample((0, 0), (0,))
ds.addSample((0, 1), (1,))
ds.addSample((1, 0), (1,))
ds.addSample((1, 1), (0,))

network = buildNetwork(2, 3, 1, bias=True, hiddenclass=TanhLayer)
trainer = BackpropTrainer(network, ds)
trainer.trainUntilConvergence()

I'm now not sure how to test this network with new data. I have tried the activate() method of the Network class ( http://pybrain.org/docs/api/structure/networks.html ) and the testOnClassData() method of the Trainer class ( http://pybrain.org/docs/api/supervised/trainers.html ), but a) I'm not sure how they work, and b) based on the documentation, I'm not sure they serve my purpose, which is to train the network to successfully predict an outcome given input parameters.

Does anyone know how to test a neural network developed in PyBrain such as mine? Thank you very much in advance! :)

Answer 1

ts = UnsupervisedDataSet(inputLength,)
ts.addSample((0,0))
net.activateOnDataset(ts)[0]

Answer 2

To test the network now feed some inputs without output.

 from pybrain.datasets import UnsupervisedDataSet  
    dst = UnsupervisedDataSet(2, )
    dst.addSample((0, 1))
    result=network.activateOnDataset(dst)[0]
    print(result)

if you use multiple inputs use for loop for

network.activateOnDataset()