Programming Neural Networks with Python

5.6    The Initialization ("__init__")

Let’s take another look at the code: we define the MLP class, which is home to the methods for building, analyzing (predict), and printing (print) the network architecture (see Listing 5.3).

In the MLP class, the __init__ method is also used to initialize objects, as in all classes. We’ve defined some self-explanatory parameters to use to influence the appearance of the network, but the basic architecture is currently hard-wired with an input layer, a hidden layer, and an output layer.

# Define the network classclass MLP(object):    """ The multilayer perceptron, MLP """     def __init__(self,                  n_input_neurons=2,n_hidden_neurons=2,n_output_neurons=1,                 weights=None,                 *args, **kwargs):        """ Initialization of the