I have a question about the NeuralNetClassifier. It’s not clear to me what the “module__n_hidden_neurons_multiplier” exactly is. I read the definition in the document, but still didn’t understand.
Is there any way we can find out the number of hidden layers while using NeuralNetClassifier?

Thank you in advance.

The neural net classifier determines the dimensionality of the dataset when you pass this dataset to fit. Let’s call this dimensionality N. The NN will have one or several hidden layers. For example, when the module__n_layers parameter is 2 there is one hidden layer. This hidden layer has a number of neurons that is module__n_hidden_neurons_multiplier x N.

1 Like