tensorflow使用神经网络实现mnist分类

本文实例为大家分享了tensorflow神经网络实现mnist分类的具体代码，供大家参考，具体内容如下

只有两层的神经网络，直接上代码

#引入包
import tensorflow as tf
import numpy as np
import matplotlib.pyplot as plt
#引入input_data文件
from tensorflow.examples.tutorials.mnist import input_data
#读取文件
mnist = input_data.read_data_sets('F:/mnist/data/',one_hot=True)

#定义第一个隐藏层和第二个隐藏层，输入层输出层
n_hidden_1 = 256
n_hidden_2 = 128
n_input = 784
n_classes = 10

#由于不知道输入图片个数，所以用placeholder
x = tf.placeholder("float",[None,n_input])
y = tf.placeholder("float",[None,n_classes])

stddev = 0.1

#定义权重
weights = {
   'w1':tf.Variable(tf.random_normal([n_input,n_hidden_1],stddev = stddev)),
   'w2':tf.Variable(tf.random_normal([n_hidden_1,n_hidden_2],stddev=stddev)),
   'out':tf.Variable(tf.random_normal([n_hidden_2,n_classes],stddev=stddev))    
   }

#定义偏置
biases = {
   'b1':tf.Variable(tf.random_normal([n_hidden_1])),
   'b2':tf.Variable(tf.random_normal([n_hidden_2])),
   'out':tf.Variable(tf.random_normal([n_classes])),
   }
print("Network is Ready")

#前向传播
def multilayer_perceptrin(_X,_weights,_biases):
 layer1 = tf.nn.sigmoid(tf.add(tf.matmul(_X,_weights['w1']),_biases['b1']))
 layer2 = tf.nn.sigmoid(tf.add(tf.matmul(layer1,_weights['w2']),_biases['b2']))
 return (tf.matmul(layer2,_weights['out'])+_biases['out'])

#定义优化函数，精准度等
pred = multilayer_perceptrin(x,weights,biases)
cost = tf.reduce_mean(tf.nn.softmax_cross_entropy_with_logits(logits = pred,labels=y))
optm = tf.train.GradientDescentOptimizer(learning_rate = 0.001).minimize(cost)
corr = tf.equal(tf.argmax(pred,1),tf.argmax(y,1))
accr = tf.reduce_mean(tf.cast(corr,"float"))
print("Functions is ready")

#定义超参数
training_epochs = 80
batch_size = 200
display_step = 4

#会话开始
init = tf.global_variables_initializer()
sess = tf.Session()
sess.run(init)

#优化
for epoch in range(training_epochs):
 avg_cost=0.
 total_batch = int(mnist.train.num_examples/batch_size)

for i in range(total_batch):
   batch_xs,batch_ys = mnist.train.next_batch(batch_size)
   feeds = {x:batch_xs,y:batch_ys}
   sess.run(optm,feed_dict = feeds)
   avg_cost += sess.run(cost,feed_dict=feeds)
 avg_cost = avg_cost/total_batch

if (epoch+1) ％ display_step ==0:
   print("Epoch:％03d/％03d cost:％.9f"％(epoch,training_epochs,avg_cost))
   feeds = {x:batch_xs,y:batch_ys}
   train_acc = sess.run(accr,feed_dict = feeds)
   print("Train accuracy:％.3f"％(train_acc))
   feeds = {x:mnist.test.images,y:mnist.test.labels}
   test_acc = sess.run(accr,feed_dict = feeds)
   print("Test accuracy:％.3f"％(test_acc))
print("Optimization Finished")

程序部分运行结果如下：

Train accuracy:0.605
Test accuracy:0.633
Epoch:071/080 cost:1.810029302
Train accuracy:0.600
Test accuracy:0.645
Epoch:075/080 cost:1.761531130
Train accuracy:0.690
Test accuracy:0.649
Epoch:079/080 cost:1.711757494
Train accuracy:0.640
Test accuracy:0.660
Optimization Finished

来源：https://blog.csdn.net/Missayaaa/article/details/80065319