mnist入门训练

'''

train_images = load_train_images()
train_labels = load_train_labels()
assert train_images.shape[0]==train_labels.shape[0],(
    "train_images.shape%s,train_labels.shape%s"%(train_images.shape,train_labels.shape)
)




test_labels = load_test_labels()
test_images = load_test_images()
assert test_images.shape[0]==test_labels.shape[0],(
    "test_images.shape%s,test_labels.shape%s"%(test_images.shape,test_labels.shape)
)


'''



import tensorflow as tf

x =tf.placeholder("float",[None,784])
W = tf.Variable(tf.zeros([784,10]))
b = tf.Variable(tf.zeros([10]))


y = tf.nn.softmax(tf.matmul(x,W)+b)



y_ = tf.placeholder("float",[None,10])
cross_entropy = -tf.reduce_sum(y_*tf.log(y))


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


init = tf.initialize_all_variables()
sess = tf.InteractiveSession()
sess.run(init)


start = 0
end = 100
now_shuffle = np.arange(60000)
np.random.shuffle(now_shuffle)
number=0
for _ in range(300):
    np.random.shuffle(now_shuffle)
    start_end = now_shuffle[start:end]
    batch_xs,batch_ys = train_images[start_end],train_labels[start_end]
    start+=100
    end+=100
    sess.run(train_step,feed_dict={x:batch_xs,y_:batch_ys})

correct_prediction = tf.equal(tf.arg_max(y,1),tf.arg_max(y_,1))
accuracy = tf.reduce_mean(tf.cast(correct_prediction,'float'))
print number,sess.run(accuracy,feed_dict={x:test_images,y_:test_labels})