本篇部落格主要介紹通過TensorFlow實現MNIST資料集的手寫數字識別。

準備資料：

首先需要獲取資料，可以通過以下程式碼進行獲取：

from tensorflow.examples.tutorials.mnist import input_data
# 獲取資料,number 1 to 10
mnist = input_data.read_data_sets('MNIST_data', one_hot=True)

注：由於使用以上程式碼獲取資料經常獲取不到，因此需要先對資料進行下載，在程式碼同目錄下建立MNIST_data目錄，並在http://yann.lecun.com/exdb/mnist/下載下面四個檔案，不用解壓直接放到MNIST_data目錄下。

搭建網路：

MNIST資料集包含了55000張訓練圖片，每張圖片的解析度為28x28，即網路的輸入為28x28=784個畫素，黑色的部分值值為1，白色的部分值為0

xs = tf.placeholder(tf.float32, [None, 784])  # 28x28

每張圖片表示一個數字，即輸出為10類，如輸出為[0 1 0 0 0 0 0 0 0 0]表示數字1

ys = tf.placeholder(tf.float32, [None, 10])

計算損失：

啟用函式選用softmax，softmax經常用於classification（分類）。

prediction = add_layer(xs, 784, 10, activation_function=tf.nn.softmax)
 

損失函式選用交叉熵函式，交叉熵函式用來衡量預測值和真實值之間的相似程度。如果完全相同，他們的交叉熵為0.

cross_entropy = tf.reduce_mean(-tf.reduce_sum(ys * tf.log(prediction),
                                              reduction_indices=[1]))

選用梯度下降演算法更新引數。

完整程式碼：

# encoding:utf-8
import tensorflow as tf
from tensorflow.examples.tutorials.mnist import input_data


# 獲取資料,number 1 to 10
mnist = input_data.read_data_sets('MNIST_data', one_hot=True)


def add_layer(inputs, in_size, out_size, activation_function=None):
    with tf.name_scope('layer'):
        with tf.name_scope('weights'):
            W = tf.Variable(tf.random_normal([in_size, out_size]), name='W')
        with tf.name_scope('bias'):
            b = tf.Variable(tf.zeros([1, out_size]) + 0.1, name='b')
        with tf.name_scope('Wx_plus_b'):
            Wx_plus_b = tf.matmul(inputs, W) + b
        if activation_function is None:
            outputs = Wx_plus_b
        else:
            outputs = activation_function(Wx_plus_b)
        return outputs


def compute_accuracy(v_xs, v_ys):
    global prediction
    y_pre = sess.run(prediction, feed_dict={xs: v_xs})
    corrct_prediction = tf.equal(tf.argmax(y_pre, 1), tf.argmax(v_ys, 1))
    accuracy = tf.reduce_mean(tf.cast(corrct_prediction, tf.float32))
    result = sess.run(accuracy, feed_dict={xs: v_xs, ys: v_ys})
    return result

# define placeholder for inputs to network
xs = tf.placeholder(tf.float32, [None, 784])  # 28x28
ys = tf.placeholder(tf.float32, [None, 10])

# add output layer, softmax通常用於做classification
prediction = add_layer(xs, 784, 10, activation_function=tf.nn.softmax)

# the error between prediction and real data
cross_entropy = tf.reduce_mean(-tf.reduce_sum(ys * tf.log(prediction),
                                              reduction_indices=[1]))

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

sess = tf.Session()

# important step
sess.run(tf.initialize_all_variables())

for i in range(1000):
    batch_xs, batch_ys = mnist.train.next_batch(100)
    sess.run(train_step, feed_dict={xs: batch_xs, ys:batch_ys})
    if i % 50 == 0:
       print(compute_accuracy(
           mnist.test.images, mnist.test.labels
       ))
 

執行結果：

Extracting MNIST_data\train-images-idx3-ubyte.gz
Extracting MNIST_data\train-labels-idx1-ubyte.gz
Extracting MNIST_data\t10k-images-idx3-ubyte.gz
Extracting MNIST_data\t10k-labels-idx1-ubyte.gz
2018-07-09 15:15:20.559165: W c:\l\tensorflow_1501918863922\work\tensorflow-1.2.1\tensorflow\core\platform\cpu_feature_guard.cc:45] The TensorFlow library wasn't compiled to use SSE instructions, but these are available on your machine and could speed up CPU computations.
2018-07-09 15:15:20.559887: W c:\l\tensorflow_1501918863922\work\tensorflow-1.2.1\tensorflow\core\platform\cpu_feature_guard.cc:45] The TensorFlow library wasn't compiled to use SSE2 instructions, but these are available on your machine and could speed up CPU computations.
2018-07-09 15:15:20.560547: W c:\l\tensorflow_1501918863922\work\tensorflow-1.2.1\tensorflow\core\platform\cpu_feature_guard.cc:45] The TensorFlow library wasn't compiled to use SSE3 instructions, but these are available on your machine and could speed up CPU computations.
2018-07-09 15:15:20.561141: W c:\l\tensorflow_1501918863922\work\tensorflow-1.2.1\tensorflow\core\platform\cpu_feature_guard.cc:45] The TensorFlow library wasn't compiled to use SSE4.1 instructions, but these are available on your machine and could speed up CPU computations.
2018-07-09 15:15:20.561767: W c:\l\tensorflow_1501918863922\work\tensorflow-1.2.1\tensorflow\core\platform\cpu_feature_guard.cc:45] The TensorFlow library wasn't compiled to use SSE4.2 instructions, but these are available on your machine and could speed up CPU computations.
2018-07-09 15:15:20.562236: W c:\l\tensorflow_1501918863922\work\tensorflow-1.2.1\tensorflow\core\platform\cpu_feature_guard.cc:45] The TensorFlow library wasn't compiled to use AVX instructions, but these are available on your machine and could speed up CPU computations.
2018-07-09 15:15:20.562993: W c:\l\tensorflow_1501918863922\work\tensorflow-1.2.1\tensorflow\core\platform\cpu_feature_guard.cc:45] The TensorFlow library wasn't compiled to use AVX2 instructions, but these are available on your machine and could speed up CPU computations.
2018-07-09 15:15:20.563277: W c:\l\tensorflow_1501918863922\work\tensorflow-1.2.1\tensorflow\core\platform\cpu_feature_guard.cc:45] The TensorFlow library wasn't compiled to use FMA instructions, but these are available on your machine and could speed up CPU computations.
WARNING:tensorflow:From D:\Users\Seavan_CC\Anaconda3\lib\site-packages\tensorflow\python\util\tf_should_use.py:170: initialize_all_variables (from tensorflow.python.ops.variables) is deprecated and will be removed after 2017-03-02.
Instructions for updating:
Use `tf.global_variables_initializer` instead.
0.0908
0.636
0.733
0.7702
0.7961
0.8105
0.824
0.8305
0.838
0.8426
0.8491
0.8514
0.8518
0.8556
0.8625
0.8645
0.8666
0.8704
0.8735
0.8699