首先我导入了numpy库：
import numpy as np
完成后，我创建了NeuralNetwork类：
class NeuralNetwork(object):
def __init__(self, input_nodes, hidden_nodes, output_nodes, learning_rate):
self.input_nodes = input_nodes
self.hidden_nodes = hidden_nodes
self.output_nodes = output_nodes
self.weights_input_to_hidden = np.random.normal(0.0, self.input_nodes ** -0.5,
(self.input_nodes,
self.hidden_nodes))
self.weights_hidden_to_output = np.random.normal(0.0, self.hidden_nodes ** -0.5,
(self.hidden_nodes,
self.output_nodes))
self.lr = learning_rate
此方法执行sigmoid操作：
def sigmoid(x):
return 1/(1+np.exp(-x))
self.activation_function=sigmoid
def train(self, features, targets):
n_records = features.shape[0]
delta_weights_i_h = np.zeros(self.weights_input_to_hidden.shape)
delta_weights_h_o = np.zeros(self.weights_hidden_to_output.shape)
for X, y in zip(features, targets):
final_outputs, hidden_outputs=self.forward_pass_train(X)
delta_weights_i_h, delta_weights_h_o = self.backpropagation(final_outputs,
hidden_outputs,
X, y,
delta_weights_i_h,
delta_weights_h_o)
self.update_weights(delta_weights_i_h, delta_weights_h_o, n_records)
def forward_pass_train(self, X):
hidden_inputs=np.dot(X, self.weights_input_to_hidden)
hidden_outputs=self.activation_function(hidden_inputs)
final_inputs=np.dot(hidden_outputs, self.weights_hidden_to_output)
final_outputs=self.activation_function(final_inputs)
return final_outputs, hidden_outputs
此方法实现了反向传播步骤，即发生错误的步骤。
def backpropagation(self, final_outputs, hidden_outputs, X, y, delta_weights_i_h, delta_weights_h_o):
'''final_outputs: Output from forward pass
y: target label
delta_weights_i_h: Change in weights from input to hidden
delta_weights_h_o: Change in weights from hidden to output
'''
error=y-final_outputs
output_error_term=error*final_outputs*(1-final_outputs)
hidden_error=np.dot(output_error_term, self.weights_hidden_to_output)
hidden_error_term=hidden_error*hidden_outputs*(1-hidden_outputs)
delta_weights_i_h+=output_error_term*hidden_outputs
delta_weights_h_o+=hidden_error_term[:,None]*hidden_outputs[:,None]
return delta_weights_i_h, delta_weights_h_o
def update_weights(self, delta_weights_i_h, delta_weights_h_o, n_records):
'''delta_weights_i_h: Change in weights from input to hidden layers
delta_weights_h_o: Change in weights from hidden to output layers
n_records: Number of records
'''
self.weights_hidden_to_output+=self.lr*delta_weights_h_o/n_records
self.weights_input_to_hidden+=self.lr*delta_weights_i_h/n_records
def run(self, features):
hidden_inputs=self.forward_pass_train(features)
hidden_outputs=self.activation_function(hidden_inputs)
final_inputs=self.forward_pass_train(hidden_outputs)
final_outputs=self.activation_function(final_inputs)
return final_outputs
iterations=100
learning_rate=0.1
hidden_nodes=2
output_nodes=1
我得到的错误看起来像这样：
ERROR: test_train (__main__.TestMethods)
----------------------------------------------------------------------
Traceback (most recent call last):
File "<ipython-input-12-b897b3b5446c>", line 18, in test_train
network.train(inputs, targets)
File "C:/Users/shivaraj/Desktop/Deep Learning/Module_1_Face_Recognition/face_recognition_nocomment.py", line 28, in train
delta_weights_i_h, delta_weights_h_o=self.backpropagation(final_outputs, hidden_outputs, X,y, delta_weights_i_h, delta_weights_h_o)
File "C:/Users/shivaraj/Desktop/Deep Learning/Module_1_Face_Recognition/face_recognition_nocomment.py", line 51, in backpropagation
hidden_error=np.dot(output_error_term, self.weights_hidden_to_output)
ValueError: shapes (1,) and (2,1) not aligned: 1 (dim 0) != 2 (dim 0)