import tensorflow as tf
from tensorflow.examples.tutorials.mnist import input_data
mnist=input_data.read_data_sets("tmp/data", one_hot=True)

learning_rate=0.01
training_epochs=25
batch_size=100
display_step=1

# placeholder   x,y 用來存儲輸入，輸入圖像x構成一個2維的浮點張量，[None,784]是簡單的平鋪圖，‘None‘代表處理的批次大小，是任意大小
x=tf.placeholder(tf.float32,[None,784])
y=tf.placeholder(tf.float32,[None,10])

# variables    為模型定義權重和偏置
w=tf.Variable(tf.zeros([784,10]))
b=tf.Variable(tf.zeros([10]))

pred=tf.nn.softmax(tf.matmul(x,w)+b)  # w*x+b要加上softmax函數

 

# reduce_sum 對所有類別求和，reduce_mean 對和取平均
cost=tf.reduce_mean(-tf.reduce_sum(y*tf.log(pred),reduction_indices=1))

# 往graph中添加新的操作，計算梯度，計算參數的更新
optimizer=tf.train.GradientDescentOptimizer(learning_rate).minimize(cost)

init=tf.initialize_all_variables()


with tf.Session() as sess:
    sess.run(init)
    for epoch in range(training_epochs):
        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)
            sess.run(optimizer,feed_dict={x:batch_xs,y:batch_ys})
        if( epoch+1)%display_step==0:
            print "cost=", sess.run(cost,feed_dict={x:batch_xs,y:batch_ys})


    prediction=tf.equal(tf.argmax(pred,1),tf.argmax(y,1))
    accuracy=tf.reduce_mean(tf.cast(prediction,tf.float32))
    print "Accuracy:" ,accuracy.eval({x:mnist.test.image,y:mnist.test.labels})

logistic 函數：

二分類問題

softmax 函數：

將k維向量壓縮成另一個k維向量，進行多分類，logistic 是softmax的一個例外

tensorflow--logistic regression