在Tensorflow中实现梯度下降法更新参数值

我就废话不多说了，直接上代码吧！

tf.train.GradientDescentOptimizer(0.01).minimize(cross_entropy)

TensorFlow经过使用梯度下降法对损失函数中的变量进行修改值，默认修改tf.Variable(tf.zeros([784,10]))

为Variable的参数。

train_step = tf.train.GradientDescentOptimizer(0.01).minimize(cross_entropy,var_list=[w,b])

也可以使用var_list参数来定义更新那些参数的值

#导入Minst数据集import input_datamnist = input_data.read_data_sets("data",one_hot=True) #导入tensorflow库import tensorflow as tf #输入变量，把28*28的图片变成一维数组（丢失结构信息）x = tf.placeholder("float",[None,784]) #权重矩阵，把28*28=784的一维输入，变成0-9这10个数字的输出w = tf.Variable(tf.zeros([784,10]))#偏置b = tf.Variable(tf.zeros([10])) #核心运算，其实就是softmax（x*w+b）y = tf.nn.softmax(tf.matmul(x,w) + b) #这个是训练集的正确结果y_ = tf.placeholder("float",[None,10]) #交叉熵，作为损失函数cross_entropy = -tf.reduce_sum(y_ * tf.log(y)) #梯度下降算法，最小化交叉熵train_step = tf.train.GradientDescentOptimizer(0.01).minimize(cross_entropy) #初始化，在run之前必须进行的init = tf.initialize_all_variables()#创建session以便运算sess = tf.Session()sess.run(init) #迭代1000次for i in range(1000): #获取训练数据集的图片输入和正确表示数字 batch_xs, batch_ys = mnist.train.next_batch(100) #运行刚才建立的梯度下降算法，x赋值为图片输入，y_赋值为正确的表示数字 sess.run(train_step,feed_dict = {x:batch_xs, y_: batch_ys}) #tf.argmax获取最大值的索引。比较运算后的结果和本身结果是否相同。#这步的结果应该是[1,1,1,1,1,1,1,1,0,1...........1,1,0,1]这种形式。#1代表正确，0代表错误correct_prediction = tf.equal(tf.argmax(y,1), tf.argmax(y_,1)) #tf.cast先将数据转换成float，防止求平均不准确。#tf.reduce_mean由于只有一个参数，就是上面那个数组的平均值。accuracy = tf.reduce_mean(tf.cast(correct_prediction,"float"))#输出print(sess.run(accuracy,feed_dict={x:mnist.test.images,y_: mnist.test.labels}))

计算结果如下

"C:/Program Files/Anaconda3/python.exe" D:/pycharmprogram/tensorflow_learn/softmax_learn/softmax_learn.pyExtracting data/train-images-idx3-ubyte.gzExtracting data/train-labels-idx1-ubyte.gzExtracting data/t10k-images-idx3-ubyte.gzExtracting data/t10k-labels-idx1-ubyte.gzWARNING:tensorflow:From C:/Program Files/Anaconda3/lib/site-packages/tensorflow/python/util/tf_should_use.py:175: 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.2018-05-14 15:49:45.866600: W C:/tf_jenkins/home/workspace/rel-win/M/windows/PY/35/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.2018-05-14 15:49:45.866600: W C:/tf_jenkins/home/workspace/rel-win/M/windows/PY/35/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.0.9163 Process finished with exit code 0