1#load mnist data
2import
tensorflow.examples.tutorials.mnist.input_data as input_data
3 mnist = input_data.read_data_sets("
mnist_data/
",one_hot=true)45
#start tensorflow interactivesession
6import
tensorflow as tf
7 sess =tf.interactivesession()89
#weight initilization
10def
weight_variable(shape):
11 initial = tf.truncated_normal(shape, stddev=0.1)
12return
tf.variable(initial)
1314
defbias_variable(shape):
15 initial = tf.constant(0.1, shape=shape)
16return
tf.variable(initial)
1718
#convolution
19def
conv2d(x, w):
20return tf.nn.conv2d(x, w, strides=[1,1,1,1], padding='
same')
2122
#pooling
23def
max_pool_2x2(x):
24return tf.nn.max_pool(x, ksize=[1,2,2,1],strides=[1,2,2,1], padding='
same')
2526
#create the model27#
placeholder
28 x = tf.placeholder("
float
",[none, 784])
29 y_ = tf.placeholder("
float
", [none, 10])
3031
#variable
32 w = tf.variable(tf.zeros([784,10]))
33 b = tf.variable(tf.zeros([10]))
3435 y = tf.nn.softmax(tf.matmul(x,w) +b)
3637
#first convolutional layer
38 w_conv1 = weight_variable([5,5,1,32])
39 b_conv1 = bias_variable([32])
4041 x_image = tf.reshape(x,[-1,28,28,1])
4243 h_conv1 =tf.nn.relu(conv2d(x_image,w_conv1) +b_conv1)
44 h_pool1 =max_pool_2x2(h_conv1)
4546
#second convolutional layer
47 w_conv2 = weight_variable([5,5,32,64])
48 b_conv2 = bias_variable([64])
4950 h_conv2 =tf.nn.relu(conv2d(h_pool1, w_conv2) +b_conv2)
51 h_pool2 =max_pool_2x2(h_conv2)
5253
#densely connected layer
54 w_fc1 = weight_variable([7*7*64, 1024])
55 b_fc1 = bias_variable([1024])
5657 h_pool2_flat = tf.reshape(h_pool2, [-1,7*7*64])
58 h_fc1 = tf.nn.relu(tf.matmul(h_pool2_flat, w_fc1) +b_fc1)
5960
#dropout
61 keep_prob = tf.placeholder("
float")
62 h_fc1_drop =tf.nn.dropout(h_fc1, keep_prob)
6364
#readout layer
65 w_fc2 = weight_variable([1024,10])
66 b_fc2 = bias_variable([10])
6768 y_conv = tf.nn.softmax(tf.matmul(h_fc1_drop,w_fc2) +b_fc2)
6970
#train and evaluate the model
71 cross_entropy = -tf.reduce_sum(y_*tf.log(y_conv))72#
train_step = tf.train.adagradoptimizer(1e-4).minimize(cross_entropy)
73 train_step = tf.train.gradientdescentoptimizer(1e-4).minimize(cross_entropy)
74 correct_prediction = tf.equal(tf.argmax(y_conv, 1), tf.argmax(y_, 1))
75 accuracy = tf.reduce_mean(tf.cast(correct_prediction, "
float"))
76sess.run(tf.initialize_all_variables())
77for i in range(5000):
78 batch = mnist.train.next_batch(50)
79if i%100 ==0:
80 train_accuracy = accuracy.eval(feed_dict=)
81print
"step %d, train accuracy %g
" %(i,train_accuracy)
82 train_step.run(feed_dict=)
8384
"test accuracy %g
" % accuracy.eval(fedd_dict=)
同樣是極客學院的課程，其實也是翻譯的國外的robot-ai部落格上的內容，但是這個部落格，現在打不開了，可能是牆的問題？沒有太深究。
tensorflow教程學習三深入MNIST
載入資料 from tensorflow.examples.tutorials.mnist import input data mnist input data.read data sets mnist data one hot true 我們使用interactivesession類可以更加靈活地...
Tensorflow 入門記錄
tensor為張量，flow為流圖。tensorflow內含有很多寫好的工具，如梯度下降演算法，卷積操作等。在使用tensorflow時，先導入包import tensorflow as tf，在進行定義tensorflow變數時，使用tf.variable 引數 有趣的是乙個叫做佔位符的工具，tf...
tensorflow入門例子
import tensorflow as tf import numpy as np 使用 numpy 生成假資料 phony data 總共 100 個點.100,2 x data np.float32 np.random.rand 100,2 隨機輸入 y data np.dot x data,...