class NeuralNetwork:
"""A fully-connected neural network."""
def __init__(self, structure):
"""Creates a neural network.
Arguments:
structure (array): Array of numbers giving the number of neurons in
each layer of the neural network.
"""
self.first_layer = Layer(structure[0])
self.last_layer = self.first_layer
self.loss = 0
for i in range(1, len(structure)):
self.last_layer = Layer(structure[i], self.last_layer)
def feedforward(self, inputs):
self.first_layer.load_data(inputs)
return self.last_layer.compute()
def train(self, inputs, targets, learning_rate=0.1):
"""Adjusts the weights and biases of the neural network.
Arguments:
inputs (array): Array of inputs that should get 'targets' as a result.
targets (array): The expected result from the inputs.
learning_rate (float): The learning rate for the step of the training.
"""
self.first_layer.load_data(inputs)
self.last_layer.compute()
last_layer_error = self.last_layer.activations - Matrix([[e] for e in targets])
self.last_layer.backpropagate(last_layer_error, learning_rate)
loss = 0
for i in range(last_layer_error.m):
loss += last_layer_error[i, 0] ** 2
loss /= last_layer_error.m
self.loss = loss
