tensorflow从入门到实践二、手写体的识别

1、数据

MNIST

2、任务

对手写数字的图片进行分类，转成 0～9 一共 10 类

3、Softmax Regression 工作原理

可以判定为某类的特征相加，然后将这些特征转化为判定是这一类的概率

4、计算流程

5、损失

交叉熵

import tensorflow as tf
from tensorflow.examples.tutorials.mnist import input_data

mnist = input_data.read_data_sets("MNUST_data/", one_hot = True)

print(mnist.train.images.shape, mnist.train.labels.shape)
print(mnist.test.images.shape, mnist.test.labels.shape)
print(mnist.validation.images.shape, mnist.validation.labels.shape)

# Extracting MNUST_data/train-images-idx3-ubyte.gz
# Extracting MNUST_data/train-labels-idx1-ubyte.gz
# Extracting MNUST_data/t10k-images-idx3-ubyte.gz
# Extracting MNUST_data/t10k-labels-idx1-ubyte.gz
# (55000, 784) (55000, 10)
# (10000, 784) (10000, 10)
# (5000, 784) (5000, 10)

sess = tf.InteractiveSession()
x = tf.placeholder(tf.float32, [None, 784])

W = tf.Variable(tf.zeros([784, 10]))
b = tf.Variable(tf.zeros([10]))

# softmax regression
y = tf.nn.softmax(tf.matmul(x, W) + b)

# loss
y_ = tf.placeholder(tf.float32, [None, 10])
cross_entropy = tf.reduce_mean(-tf.reduce_sum(y_ * tf.log(y), reduction_indices = [1]))

# SGD 
train_step = tf.train.GradientDescentOptimizer(0.5).minimize(cross_entropy)

# 全局参数初始化
tf.global_variables_initializer().run()

# train
for i in range(1000):
    # 每次抽取100条数据
    batch_xs, batch_ys = mnist.train.next_batch(100)
    train_step.run({x:batch_xs, y_ : batch_ys})

# 返回分类是否正确的类别    
correct_prediction = tf.equal(tf.argmax(y, 1), tf.argmax(y_, 1))

# cast
accuracy = tf.reduce_mean(tf.cast(correct_prediction, tf.float32))

print(accuracy.eval({x : mnist.test.images, y_ : mnist.test.labels}))


# output 0.9211
     
     
     
     

      
      
      
      
1
      
      
      
      
2
      
      
      
      
3
      
      
      
      
4
      
      
      
      
5
      
      
      
      
6
      
      
      
      
7
      
      
      
      
8
      
      
      
      
9
      
      
      
      
10
      
      
      
      
11
      
      
      
      
12
      
      
      
      
13
      
      
      
      
14
      
      
      
      
15
      
      
      
      
16
      
      
      
      
17
      
      
      
      
18
      
      
      
      
19
      
      
      
      
20
      
      
      
      
21
      
      
      
      
22
      
      
      
      
23
      
      
      
      
24
      
      
      
      
25
      
      
      
      
26
      
      
      
      
27
      
      
      
      
28
      
      
      
      
29
      
      
      
      
30
      
      
      
      
31
      
      
      
      
32
      
      
      
      
33
      
      
      
      
34
      
      
      
      
35
      
      
      
      
36
      
      
      
      
37
      
      
      
      
38
      
      
      
      
39
      
      
      
      
40
      
      
      
      
41
      
      
      
      
42
      
      
      
      
43
      
      
      
      
44
      
      
      
      
45
      
      
      
      
46
      
      
      
      
47
      
      
      
      
48
      
      
      
      
49
      
      
      
      
50
      
      
      
      
51
      
      
      
      
52