Tensorflow hello world<Ò»>

1¡¢ÏÂÔØmnistÊý¾Ý¼¯(Èý´óÌØÕ÷)

1.1 °üº¬ÑµÁ·¼¯(55000)¡¢ÑéÖ¤¼¯(5000)ºÍ²âÊÔ¼¯(10000)

1.2 ÿÕÅͼƬ¶¼ÊÇ28*28άµÄÏòÁ¿

1.3 labelÊǸö10άµÄÈȱàÂëÏòÁ¿(one_hot)

Êý¾Ý¼¯shape´òÓ¡½á¹ûÈçÏÂËùʾ

(55000, 784)

(55000, 10)

(5000, 784)

(5000, 10)

(10000, 784)

(10000, 10)

2¡¢½¨Á¢·ÖÀàÄ£Ðͼ°¼ì²âÄ£ÐÍ׼ȷ¶È

ʹÓÃÇ°¼¸Ìì½²¹ýµÄTensorflow»ù´¡½¨Á¢³õ²½Ä£ÐÍ~

fromtensorflow.examples.tutorials.mnistimportinput_data

importtensorflowastf

#ÏÂÔØÊý¾Ý¼¯

mnist=input_data.read_data_sets("mnist_data/",one_hot=True)

#¶¨ÒåÊäÈëÊý¾Ýռλ·û

train_input=tf.placeholder(tf.float32,[None,784])

train_label=tf.placeholder(tf.float32,[None,10])

#¶¨Òå³õʼ»¯²ÎÊý

W=tf.Variable(tf.random_normal([784,10],stddev=0.1))

b=tf.Variable(tf.zeros([10]))

#¶¨ÒåѵÁ·Ä£ÐÍ

pred_y=tf.nn.softmax(tf.matmul(train_input,W)+b)

#¶¨ÒåËðʧº¯Êý

cross_entropy=tf.reduce_mean(-tf.reduce_sum(train_label*tf.log(pred_y)))

#¶¨ÒåËðʧº¯ÊýµÄÓÅ»¯·½·¨

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

#½¨Á¢»á»°

sess=tf.Session()

#³õʼ»¯±äÁ¿

sess.run(tf.global_variables_initializer())

#ѵÁ·Êý¾Ý¼¯

foriinrange(1000):

batch_xs,batch_ys=mnist.train.next_batch(128)

sess.run(train_step,feed_dict={train_input:batch_xs,train_label:batch_ys})

#¼ÆËãÄ£ÐÍ׼ȷÐÔ

correct_pre=tf.equal(tf.argmax(train_label,1),tf.argmax(pred_y,1))

acc=tf.reduce_mean(tf.cast(correct_pre,tf.float32))

print(sess.run(acc,feed_dict={train_input:mnist.test.images,train_label:mnist.test.labels}))

׼ȷÂÊ91%×óÓÒ

¿Îºó×÷Òµ£º¾í»ýÉñ¾­ÍøÂçµÄÊýѧ»ù´¡

¶¯ÊÖÊÔÒ»ÊÔÉÏÃæµÄСʾÀý£¬Ã¿Ììѧϰһµãµã~