MNIST数据集两层神经网络分类
MNIST数据集两层神经网络分类
>>> import tensorflow as tf
>>> from tensorflow.examples.tutorials.mnist import input_data
>>> mnist=input_data.read_data_sets("MNIST_data/",one_hot=True)
//导入数据集到对象mnist
>>> x=tf.placeholder(tf.float32,[None,784])//训练图像的占位符
>>> y_=tf.placeholder(tf.float32,[None,10])//训练标签的占位符
>>> x_image=tf.reshape(x,[-1,28,28,1])
//将单张图片从784维转出28*28矩阵图片
>>> def weight_variable(shape):
initial=tf.truncated_normal(shape,stddev=0.1)
return tf.Variable(initial)
//创建卷积核,
>>> def bias_variable(shape):
initial=tf.constant(0.1,shape=shape)
return tf.Variable(initial)
//创建偏置项
>>> def conv2d(x,W):
return tf.nn.conv2d(x,W,strides=[1,1,1,1],padding='SAME')
//卷积计算
>>> def max_pool_2x2(x):
return tf.nn.max_pool(x,ksize=[1,2,2,1],strides=[1,2,2,1],padding='SAME')
//最大池化
#第一层卷积层
>>> W_conv1=weight_variable([5,5,1,32])//卷积核
>>> b_conv1=bias_variable([32])//偏置项
>>> h_conv1=tf.nn.relu(conv2d(x_image,W_conv1)+b_conv1)
>//卷积计算,并用relu作为激活函数
>>> h_pool1=max_pool_2x2(h_conv1)
//将上一步卷积结果池化
#第二层卷积层
>>> W_conv2=weight_variable([5,5,32,64])
>>> b_conv2=bias_variable([64])
>>> h_conv2=tf.nn.relu(conv2d(h_pool1,W_conv2)+b_conv2)
>>> h_pool2=max_pool_2x2(h_conv2)
#全连接层
>>> W_fc1=weight_variable([7*7*64,1024])
>>> b_fc1=bias_variable([1024])
>>> 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)
>>> keep_prob=tf.placeholder(tf.float32)
>>> h_fc1_drop=tf.nn.dropout(h_fc1,keep_prob)
//使用Dropout,keep_prob是一个占位符,训练时为0.5,测试时为1
在全连接层中加入Dropout,它是防止神经网络过拟合的一种手段,在每一步训练的时候,以一定概率去掉网络中的某些连接,但这不是永久的,只在当前步骤中去除,且每一次是随机去除的。
>>> W_fc2=weight_variable([1024,10])
>>> b_fc2=bias_variable([10])
>>> y_conv=tf.matmul(h_fc1_drop,W_fc2)+b_fc2
//再加入一层全连接层把上一步得到的h_fc1_drop转化为10个类别的打分
>>> cross_entropy=tf.reduce_mean(
tf.nn.softmax_cross_entropy_with_logits(labels=y_,logits=y_conv))
//不采用先softmax再计算交叉熵的方法
而是直接利用tf.nn.softmax_cross_entropy_with_logits计算交叉熵
>>> train_step=tf.train.AdamOptimizer(1e-4).minimize(cross_entropy)
//采用Adam优化交叉熵
>>> correct_prediction=tf.equal(tf.argmax(y_conv,1),tf.argmax(y_,1))
>>> accuracy=tf.reduce_mean(tf.cast(correct_prediction,tf.float32))
//定义测试的准确率
>>> sess=tf.InteractiveSession()//创建会话
>>> sess.run(tf.global_variables_initializer())//对变量初始化
>>> for i in range(20000):
batch=mnist.train.next_batch(50)
//训练20000次,每次batch为50个样本
#每100次报告在验证集上的准确度
>>> 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 %g" % (i,train_accuracy))
>>> train_step.run(feed_dict={x:batch[0],y_:batch[1],keep_prob:0.5})
//训练
>>> print("test accuracy %g"%accuracy.eval(feed_dict={
x:mnist.test.images, y_:mnist.test.labels,keep_prob:1.0}))
//报告测试集上的准确率