自学Tensorflow之交叉熵MNIST分类

本文用于复习TF官网教学内容

第一节MNIST初学

主要复习内容为交叉熵、梯度下降

tensorflow1.0 mac osx 10.12.4

---

• import部分

from __future__ import absolute_import
from __future__ import division
from __future__ import print_function

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

• 利用placeholder（占位符）构图

# data
mnist = input_data.read_data_sets('/tmp/tensorflow/mnist/input_data',one_hot=True)
# placeholder for x cause image=28*28
x= tf.placeholder(tf.float32,[None,784])
# predict y
W = tf.Variable(tf.zeros([784,10]))
b = tf.Variable(tf.zeros([10]))
y = tf.nn.softmax(tf.matmul(x,W)+b)

# correct y
y_ = tf.placeholder(tf.float32,[None, 10])

• 计算交叉熵，并用梯度下降进行训练

# calculate cross_entropy
cross_entropy = tf.reduce_mean(tf.nn.softmax_cross_entropy_with_logits(labels=y_, logits=y))
# use GradientDescent for training
train_step = tf.train.GradientDescentOptimizer(0.5).minimize(cross_entropy)

• 启动session（交互或with tf.Session() as sess）并初始化变量

sess = tf.InteractiveSession()
tf.global_variables_initializer().run()

• 训练100次，每次（一个batch）100个，mnist自带next_batch方法；把数据喂给train_step

for _ in range(100):
    batch_xs, batch_ys = mnist.train.next_batch(100)
    sess.run(train_step,feed_dict={x:batch_xs,y_:batch_ys})

• y = n*10，argmax（y,1）即可把10列中最大的找出来，即是分类结果；accuracy算出correct_prediction均值，把test的数据喂给accuracy

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}))