Python神经网络实现手写数字识别

利用Python制作简易的神经网络

* 实现手写数字的识别
* github项目地址：https://github.com/coding-cheng/mnist 
* 完整的数据集：
	训练集 ：http://www.pjreddie.com/midea/files/mnist_train.csv
	测试集 ：http://www.pjreddie.com/midea/files/mnist_test.csv
* 参考书籍 : 《Python神经网络编程》
* github项目分为以下几个部分：
		1. neural_network.py : 神经网络
		2. train.py : 训练模型，并保存模型
		3. test.py : 测试模型，并且保存模型的准确率
		4. useyournum.py : 可使用自己写的数字作为测试对象
		5. mnist_dataset : 部分训练测试数据
		6. save_model : 保存训练后的模型数据
		7. yournum : 用于保存你所写的数字
		8. accuracy : 保存测试准确率

数据集网盘下载地址:
链接：https://pan.baidu.com/s/1d_JpNKqIoK86a7pcGoB_Hw 
提取码：qbn9 

• 项目结构
  

1. 简易神经网络的搭建

import numpy as np
import scipy.special

class neuralNetwork():
    # initialise the neuralNetwork
    def __init__(self, inputnodes, hiddennodes, outputnodes, learningrate):
        # set number of nodes in each input, hidden, output layer
        self.inodes = inputnodes
        self.hnodes = hiddennodes
        self.onodes = outputnodes

        # link weight matrices, wih and who
        # weights inside the arrays are w_i_j, where link is from node i to node j in the next layer
        # w11 w21
        # w12 w22 etc
        self.wih = np.random.normal(0.0, pow(self.hnodes, -0.5), (self.hnodes, self.inodes))
        self.who = np.random.normal(0.0, pow(self.onodes, -0.5), (self.onodes, self.hnodes))

        # learning rate
        self.lr = learningrate

        # activation function is the sigmoid function
        self.activation_function = lambda x: scipy.special.expit(x)

    # train the neuralNetwork
    def train(self, inputs_list, targets_list):
        # convert inputs list to 2d array
        inputs = np.array(inputs_list, ndmin=2).T
        targets = np.array(targets_list, ndmin=2).T

        # calculate signals into hidden layer
        hidden_inputs = np.dot(self.wih, inputs)
        # calculate the signals emerging from hidden layer
        hidden_outputs = self.activation_function(hidden_inputs)

        # calculate signals from final output layer
        final_inputs = np.dot(self.who, hidden_outputs)
        # calculate the signals emerging from final output layer
        final_outputs = self.activation_function(final_inputs)
        # output layer error is the (target - actual)
        output_errors = targets - final_outputs
        # hidden layer error is the output_errors, split by weights, recombined at hidden nodes
        hidden_errors = np.dot(self.who.T, output_errors)

        # update the weights for the links between the hidden and output layer
        self.who += self.lr * np.dot((output_errors * final_outputs * (1.0 - final_outputs)),
                                     np.transpose(hidden_outputs))

        # update the weights for the links between the input and hidden layer
        self.wih += self.lr * np.dot((hidden_errors * hidden_outputs * (1.0 - hidden_outputs)), np.transpose(inputs))

    # query the neural network
    def query(self, inputs_list):
        # convert inputs list to 2d array
        inputs = np.array(inputs_list, ndmin=2).T

        # calculate signals into hidden layer
        hidden_inputs = np.dot(self.wih, inputs)
        # calculate the signals emerging from hidden layer
        hidden_outputs = self.activation_function(hidden_inputs)

        # calculate signals into final output layer
        final_inputs = np.dot(self.who, hidden_outputs)
        # calculate the signals emerging from final output layer
        fianl_outputs = self.activation_function(final_inputs)

        return fianl_outputs

2. 训练模块

from neural_network import neuralNetwork
import numpy as np


# number of input, hidden and output nodes
input_nodes = 784  # 28*28
hidden_nodes = 500
output_nodes = 10

# learning rate is 0.3
learning_rate = 0.1

# create instance of neural network
n = neuralNetwork(input_nodes, hidden_nodes, output_nodes, learning_rate)

training_data_file = open("mnist_dataset/mnist_train_100.csv", "r")
training_data_list = training_data_file.readlines()
training_data_file.close()

# train the neural network

# epochs is the number of times the training data set is used for training
epochs = 5

for e in range(epochs):
    # go through all records in the training data set
    for record in training_data_list:
        # split the record by ',' commas
        all_values = record.split(',')
        # scale and shift the inputs
        inputs = (np.asfarray(all_values[1:]) / 255.0 * 0.99) + 0.01
        # create the target output values (all 0.01, expert the desired label which is 0.99)
        targets = np.zeros(output_nodes) + 0.01
        # all_values[0] is the target label for this record
        targets[int(all_values[0])] = 0.99
        n.train(inputs, targets)

save_wih_file = "save_model/wih.npy"
np.save(save_wih_file, n.wih)
save_who_file = "save_model/who.npy"
np.save(save_who_file, n.who)

3. 测试模块

from neural_network import neuralNetwork
import numpy as np

# number of input, hidden and output nodes
input_nodes = 784  # 28*28
hidden_nodes = 100
output_nodes = 10

# learning rate is 0.3
learning_rate = 0.3

# create instance of neural network
n = neuralNetwork(input_nodes, hidden_nodes, output_nodes, learning_rate)

save_wih_file = "save_model/wih.npy"
save_who_file = "save_model/who.npy"
n.wih = np.load(save_wih_file)
n.who = np.load(save_who_file)

# load the mnist test data csv file into a list
test_data_file = open("mnist_dataset/mnist_test_10.csv", "r")
test_data_list = test_data_file.readlines()
test_data_file.close()

# test the neural network
# scorecard for how well the network performs, initially empty
scorecard = []
# go through all records in the test data set
for record in test_data_list:
    # split the record by ',' commas
    all_values = record.split(',')
    # correct answer is first value
    correct_label = int(all_values[0])
    # print(correct_label, "correct_label")
    # scale and shift the inputs
    inputs = (np.asfarray(all_values[1:]) / 255.0 * 0.99) + 0.01
    # query the network
    outputs = n.query(inputs)
    # the index of the highest value corresponds to the label
    label = np.argmax(outputs)
    # print(label, "network's answer")
    # append correct or incorrect to list
    if label == correct_label:
        # network's answer matches correct answer, add 1 to scorecard
        scorecard.append(1)
    else:
        # network's answer doesn't match correct answer, add 1 to scorecard
        scorecard.append(0)

# calculate the performance score, the fraction of correct answers
scorecard_array = np.asarray(scorecard)
performance = scorecard_array.sum() / scorecard_array.size
print("performance = ", performance)

save_accuracy = "accuracy.txt"
with open(save_accuracy, 'a') as f:
    f.write("Learning rate is : " + str(n.lr))
    f.write("       Accuracy is : " + str(performance) + '\n')

4. 使用你自己的手写数字

from PIL import Image
import numpy as np
from neural_network import neuralNetwork
import matplotlib.pyplot as plt
import imageio

# number of input, hidden and output nodes
input_nodes = 784  # 28*28
hidden_nodes = 100
output_nodes = 10

# learning rate is 0.3
learning_rate = 0.1

# create instance of neural network
n = neuralNetwork(input_nodes, hidden_nodes, output_nodes, learning_rate)

# change pixel value
file_in = 'yournum/yournum.png'
width = 28
height = 28
file_out = 'yournum/yournum_c.png'
image = Image.open(file_in)
resized_image = image.resize((width, height), Image.ANTIALIAS)
resized_image.save(file_out)

save_wih_file = "save_model/wih.npy"
save_who_file = "save_model/who.npy"
n.wih = np.load(save_wih_file)
n.who = np.load(save_who_file)

# test the neural network

image_file_name = "yournum/yournum_c.png"
img_array = imageio.imread(image_file_name, as_gray=True)

img_data = 255.0 - img_array.reshape(784)
img_data = (img_data / 255.0 * 0.99) + 0.01
# query the network
outputs = n.query(img_data)
for i in range(10):
    print(i, " ", outputs[i])
result = np.argmax(outputs)
print("Neural network predicts ", result)
image_array = np.asfarray(img_data).reshape((28, 28))
plt.imshow(image_array, cmap='Greys', interpolation='None')
plt.show()