一、安装

网好的直接sudo pip install tensorflow网不好的从github上下载对应版本的tensorflow，格式为whl。我下的是0.12.head从pip安装 sudo pip install /下载的位置/下载的.whl检查是否安装好，以及查看安装版本,终端命令： python import tensorflow as tf tf.__version__ 如果Import 报错了，就看看是哪个文件报错。比如是含有关键字’protobuf’的，就卸载protobuf,再重装一下。 sudo apt-get remove python-protobuf sudo pip uninstall protobuf sudo pip install protobuf 然后再重装tensorflowtensorflow 默认安装路径： /usr/local/lib/python2.7/dist-packages/tensorflow/examples/tutorials/mnist

二、mnist讲解 适合入门

MNIST机器学习入门 （http://wiki.jikexueyuan.com/project/tensorflow-zh/tutorials/mnist_beginners.html）

mnist NN分类

#coding=utf-8 from __future__ import absolute_import from __future__ import division from __future__ import print_function # Import data from tensorflow.examples.tutorials.mnist import input_data import tensorflow as tf flags = tf.app.flags FLAGS = flags.FLAGS flags.DEFINE_string('data_dir', '/media/zhangsp/softwares/tensorflowUbuntu/mnist/data/', 'Directory for storing data') mnist = input_data.read_data_sets(FLAGS.data_dir, one_hot=True) sess = tf.InteractiveSession() # create model x = tf.placeholder(tf.float32, [None, 784]) #None的意思是不预先定义行数，可以随意改变大小，28*28的像素共784个 W = tf.Variable(tf.zeros([784, 10])) #权值矩阵初始化 b = tf.Variable(tf.zeros([10])) #偏置 y = tf.nn.softmax(tf.matmul(x, W) + b) #定义y，就是线性结构的神经网络了。加权变换+softmax回归 y_ = tf.placeholder(tf.float32, [None, 10]) #定义训练标签 cross_entropy = tf.reduce_mean(-tf.reduce_sum(y_ * tf.log(y), reduction_indices=[1]))#Loss function：交叉熵 learning_rate = 0.5 train_step = tf.train.GradientDescentOptimizer(learning_rate).minimize(cross_entropy) # Train init = tf.initialize_all_variables() sess = tf.Session() sess.run(init) print ("learning rate : "+str(learning_rate)) for i in range(1000): batch_xs, batch_ys = mnist.train.next_batch(100) sess.run(train_step, feed_dict={x: batch_xs, y_: batch_ys}) if i % 100 == 0: #每100代输出一下预测结果 # Test trained model correct_prediction = tf.equal(tf.argmax(y, 1), tf.argmax(y_, 1)) accuracy = tf.reduce_mean(tf.cast(correct_prediction, tf.float32)) print(sess.run(accuracy, feed_dict={x: mnist.test.images, y_: mnist.test.labels})) #Train again learning_rate = 0.01 train_step = tf.train.GradientDescentOptimizer(learning_rate).minimize(cross_entropy) init = tf.initialize_all_variables() sess = tf.Session() sess.run(init) print ("learning rate : "+str(learning_rate)) for i in range(1000): batch_xs, batch_ys = mnist.train.next_batch(100) sess.run(train_step, feed_dict={x: batch_xs, y_: batch_ys}) if i % 100 == 0: # 每100代输出一下预测结果 # Test trained model correct_prediction = tf.equal(tf.argmax(y, 1), tf.argmax(y_, 1)) accuracy = tf.reduce_mean(tf.cast(correct_prediction, tf.float32)) print(sess.run(accuracy, feed_dict={x: mnist.test.images, y_: mnist.test.labels}))

三、 查看可视化结果

终端输入 tensorboard --logdir='logs/'可以得到一个地址，浏览器中打开这个网址就可以看可视化结果了。

四、 Mnist CNN 分类

TensorFlow学习笔记2：构建CNN模型 博客写的非常好。 （http://www.jeyzhang.com/tensorflow-learning-notes-2.html） 代码：

#coding=utf-8 import tensorflow as tf from tensorflow.examples.tutorials.mnist import input_data def weight_variable(shape): initial = tf.truncated_normal(shape, stddev=0.1)#为了避免网络对称性，初始化时加入小的噪声 return tf.Variable(initial) def bias_variable(shape): initial = tf.constant(0.1, shape=shape) return tf.Variable(initial) def conv2d(x, W): #卷积，滑动窗口步长1，padding='SAME'表示通过填充0，使得输入和输出的形状一致。 return tf.nn.conv2d(x, W, strides=[1, 1, 1, 1], padding='SAME') def max_pool_2x2(x): #池化：最大池化，2*2窗口 return tf.nn.max_pool(x, ksize=[1, 2, 2, 1], strides=[1, 2, 2, 1], padding='SAME') #读入数据和初始化 flags = tf.app.flags FLAGS = flags.FLAGS flags.DEFINE_string('data_dir', '/media/zhangsp/softwares/tensorflowUbuntu/mnist/data/', 'Directory for storing data') mnist = input_data.read_data_sets(FLAGS.data_dir, one_hot=True) sess = tf.InteractiveSession() x = tf.placeholder("float", shape=[None, 784]) y_ = tf.placeholder("float", shape=[None, 10]) W = tf.Variable(tf.zeros([784, 10])) b = tf.Variable(tf.zeros([10])) W_conv1 = weight_variable([5, 5, 1, 32]) #第一个卷积层，5×5的patch,输入1通道，输出32通道（32个特征映射/卷积核） b_conv1 = bias_variable([32]) x_image = tf.reshape(x, [-1, 28, 28, 1]) #reshape成4维，第2、3维是图片的宽和高，第四维是通道数 h_conv1 = tf.nn.relu(conv2d(x_image, W_conv1) + b_conv1)#卷积+偏置，再relu激活 h_pool1 = max_pool_2x2(h_conv1) #池化 W_conv2 = weight_variable([5, 5, 32, 64]) #第二个卷积层，提取64个特征，每个7*7大小 b_conv2 = bias_variable([64]) h_conv2 = tf.nn.relu(conv2d(h_pool1, W_conv2) + b_conv2) h_pool2 = max_pool_2x2(h_conv2) # Now image size is reduced to 7*7 W_fc1 = weight_variable([7 * 7 * 64, 1024]) #全连接层 b_fc1 = bias_variable([1024]) h_pool2_flat = tf.reshape(h_pool2, [-1, 7 * 7 * 64])#reshape成向量 h_fc1 = tf.nn.relu(tf.matmul(h_pool2_flat, W_fc1) + b_fc1)#向量乘法，h*w+b keep_prob = tf.placeholder("float") #不被droupout丢弃的概率 h_fc1_drop = tf.nn.dropout(h_fc1, keep_prob) W_fc2 = weight_variable([1024, 10]) b_fc2 = bias_variable([10]) y_conv = tf.nn.softmax(tf.matmul(h_fc1_drop, W_fc2) + b_fc2)#softmax输出 cross_entropy = -tf.reduce_sum(y_ * tf.log(y_conv)) train_step = tf.train.AdamOptimizer(1e-4).minimize(cross_entropy)#没用梯度下降哟 correct_prediction = tf.equal(tf.argmax(y_conv, 1), tf.argmax(y_, 1)) accuracy = tf.reduce_mean(tf.cast(correct_prediction, "float")) sess.run(tf.initialize_all_variables()) for i in range(2000): #我减少了一下训练量。。。本来是20000 batch = mnist.train.next_batch(50) if i % 100 == 0: train_accuracy = accuracy.eval(feed_dict={ x: batch[0], y_: batch[1], keep_prob: 1.0}) print "step %d, training accuracy %.3f" % (i, train_accuracy) train_step.run(feed_dict={x: batch[0], y_: batch[1], keep_prob: 0.5}) print "Training finished" print "test accuracy %.3f" % accuracy.eval(feed_dict={ x: mnist.test.images, y_: mnist.test.labels, keep_prob: 1.0})