2
$\begingroup$

We have trained a Neural Network on Mathemetica classify with one hidden layer and received good classification results. We would like to know the default parameters beside it has one hidden layer? Such as network architecture(fully-connected or convolution), optimizer,active function,learning rate..

$\endgroup$
1
$\begingroup$

Lets define a Classifer that uses a Neural Net as a method with some training data.

train = {{1, 2, 3} -> "Red", {2, 6, 2} -> "Blue", {4, 9, 2} -> 
  "Green"}
classModel=Classify[train, Method -> "NeuralNetwork", 
 TargetDevice -> "GPU"];

We can then Use Keys[] and Values[] to explore the options that Classify[] uses.

Keys[classModel[[1, "Model"]]]
Values[classModel[[1, "Model"]]]

We can specify the network, training and even the options to view the learning rate, optimiser method etc

ClassNN = classModel[[1, "Model", "Network"]]
TrainingNN = classModel[[1, "Model", "Training"]]
OptionsNN = classModel[[1, "Model", "Options"]]
$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.