一, 数据准备
实验数据使用 MNIST 数据集.
MNIST 数据集已经是一个被 "嚼烂" 了的数据集, 很多教程都会对它 "下手", 几乎成为一个 "典范".
在很多 tensorflow 教程中, 用下面这一句下载 mnist 数据集:
mnist = input_data.read_data_sets('MNIST_data', one_hot=True)
但实际运行时根本无法通过网络下载, 解决方案就是手工下载数据, 然后直接导入使用.
下载地址: http://yann.lecun.com/exdb/mnist/
4 个文件, 注意下载后不需要解压.
如果把上述下载的文件放在与运行的. py 文件同一个目录下, 那么导入数据的代码是这样的:
mnist = input_data.read_data_sets('./', one_hot=True)
二, 代码
- import tensorflow as tf
- from tensorflow.examples.tutorials.mnist import input_data
- # number 1 to 10 data
- mnist = input_data.read_data_sets('./', one_hot=True)
- def compute_accuracy(v_xs, v_ys):
- global prediction
- y_pre = sess.run(prediction, feed_dict={xs: v_xs, keep_prob: 1})
- correct_prediction = tf.equal(tf.argmax(y_pre,1), tf.argmax(v_ys,1))
- accuracy = tf.reduce_mean(tf.cast(correct_prediction, tf.float32))
- result = sess.run(accuracy, feed_dict={xs: v_xs, ys: v_ys, keep_prob: 1})
- return result
- # 产生随机变量, 符合 normal 分布
- # 传递 shape 就可以返回 weight 和 bias 的变量
- def weight_variable(shape):
- initial = tf.truncated_normal(shape, stddev=0.1)
- return tf.Variable(initial)
- def bias_variable(shape):
- initial = tf.constant(0.1, shape=shape)
- return tf.Variable(initial)
- # 定义 2 维的 convolutional 图层
- def conv2d(x, W):
- # stride [1, x_movement, y_movement, 1]
- # Must have strides[0] = strides[3] = 1
- # strides 就是跨多大步抽取信息
- return tf.nn.conv2d(x, W, strides=[1, 1, 1, 1], padding='SAME')
- # 定义 pooling 图层
- def max_pool_2x2(x):
- # stride [1, x_movement, y_movement, 1]
- # 用 pooling 对付跨步大丢失信息问题
- return tf.nn.max_pool(x, ksize=[1,2,2,1], strides=[1,2,2,1], padding='SAME')
- # define placeholder for inputs to network
- xs = tf.placeholder(tf.float32, [None, 784]) # 784=28x28
- ys = tf.placeholder(tf.float32, [None, 10])
- keep_prob = tf.placeholder(tf.float32)
- x_image = tf.reshape(xs, [-1, 28, 28, 1]) # 最后一个 1 表示数据是黑白的
- # print(x_image.shape) # [n_samples, 28,28,1]
- ## 1. conv1 layer ##
- # 把 x_image 的厚度 1 加厚变成了 32
- W_conv1 = weight_variable([5, 5, 1, 32]) # patch 5x5, in size 1, out size 32
- b_conv1 = bias_variable([32])
- # 构建第一个 convolutional 层, 外面再加一个非线性化的处理 relu
- h_conv1 = tf.nn.relu(conv2d(x_image, W_conv1) + b_conv1) # output size 28x28x32
- # 经过 pooling 后, 长宽缩小为 14x14
- h_pool1 = max_pool_2x2(h_conv1) # output size 14x14x32
- ## 2. conv2 layer ##
- # 把厚度 32 加厚变成了 64
- W_conv2 = weight_variable([5,5, 32, 64]) # patch 5x5, in size 32, out size 64
- b_conv2 = bias_variable([64])
- # 构建第二个 convolutional 层
- h_conv2 = tf.nn.relu(conv2d(h_pool1, W_conv2) + b_conv2) # output size 14x14x64
- # 经过 pooling 后, 长宽缩小为 7x7
- h_pool2 = max_pool_2x2(h_conv2) # output size 7x7x64
- ## 3. func1 layer ##
- # 飞的更高变成 1024
- W_fc1 = weight_variable([7*7*64, 1024])
- b_fc1 = bias_variable([1024])
- # [n_samples, 7, 7, 64] ->> [n_samples, 7*7*64]
- # 把 pooling 后的结果变平
- h_pool2_flat = tf.reshape(h_pool2, [-1, 7*7*64])
- h_fc1 = tf.nn.relu(tf.matmul(h_pool2_flat, W_fc1) + b_fc1)
- h_fc1_drop = tf.nn.dropout(h_fc1, keep_prob)
- ## 4. func2 layer ##
- # 最后一层, 输入 1024, 输出 size 10, 用 softmax 计算概率进行分类的处理
- W_fc2 = weight_variable([1024, 10])
- b_fc2 = bias_variable([10])
- prediction = tf.nn.softmax(tf.matmul(h_fc1_drop, W_fc2) + b_fc2)
- # the error between prediction and real data
- cross_entropy = tf.reduce_mean(-tf.reduce_sum(ys * tf.log(prediction),
- reduction_indices=[1])) # loss
- train_step = tf.train.AdamOptimizer(1e-4).minimize(cross_entropy)
- sess = tf.Session()
- # important step
- sess.run(tf.global_variables_initializer())
- 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, keep_prob: 0.5})
- if i % 50 == 0:
- print(compute_accuracy(
mnist.test.images, mnist.test.labels))
三, Github 代码下载
下载 https://github.com/zhenghaishu/MachineLearning/tree/master/CNN
四, 参考
- http://v.youku.com/v_show/id_XMTYyMTUyMjc0OA==.html?spm=a2hzp.8253869.0.0
- https://github.com/MorvanZhou/tutorials/tree/master/tensorflowTUT/tf18_CNN3
- https://www.jianshu.com/p/e2f62043d02b
- https://blog.csdn.net/i8088/article/details/79126150
来源: http://www.jianshu.com/p/618d4b789dd1