线性回归

import tensorflow as tf

#初始化变量和模型参数， 定义训练闭环中运算
W = tf.Variable(tf.zeros([2, 1]), name = "weights")
b = tf.Variable(0.0, name = "bias")

#计算推断模型在数据X上的输出， 并将结果返回
def inference(X):
    return tf.matmul(X, W) + b

#依据数据X及其期望输出Y的计算损失
def loss(X, Y):
    Y_predicted = inference(X)
    return tf.reduce_sum(tf.squared_difference(Y, Y_predicted))
'''
a = tf.to_float([1,2,3,4,5])
b = tf.to_float([1,1,1,1,1])
with tf.Session() as sess:
    print sess.run(tf.squared_difference(a, b))
    print sess.run(tf.square(a - b))

输出结果：
[  0.   1.   4.   9.  16.]
[  0.   1.   4.   9.  16.]

'''

#读取或生成训练数据X及其期望输出Y
def inputs():
    weights_age = [[84, 46], [73, 20], [65, 52], [70, 30], [76, 57], [69, 25],[63, 28], [72,36],
                               [79, 57], [75, 44], [27, 44], [89,31], [65, 52],[57,23], [59, 60],[69, 48], [60, 34],
                                [79, 51], [75, 50], [82, 34], [59,46], [67,23], [85, 37],[55, 40], [63, 30]
                  ]
    blood_fat_content = [ 354, 190, 405, 263, 451, 302, 288, 385, 402, 365, 209, 290, 346,
                                       254, 395, 434, 220, 374, 308, 220, 311, 181, 274, 303, 244
                        ]

    return tf.to_float(weights_age), tf.to_float(blood_fat_content)

#对训练得到的模型进行评估
def evaluate(sess, X, Y):

    print sess.run(inference([[80., 25.]]))
    print sess.run(inference([[65., 25.]]))

#依据计算的总损失训练或调整模型参数
def train(tol_loss):
    learning_rate = 0.0000001
    return tf.train.GradientDescentOptimizer(learning_rate).minimize(tol_loss)
#在一个会话中对象启动数据流图， 搭建流程
with tf.Session() as sess:

    tf.global_variables_initializer().run()
    X, Y = inputs()  

    tol_loss = loss(X, Y)
    train_op = train(tol_loss)
    #搭建多线程
    coord = tf.train.Coordinator()
    threads = tf.train.start_queue_runners()
    #实际的训练次数
    training_steps = 1000

    for step in range(training_steps):
        sess.run([train_op])
        #处于调试和学习的目的， 查看损失在训练过程中递减情况
        if step % 10 == 0:
            print "loss: ", sess.run([tol_loss])

    evaluate(sess, X, Y)
    coord.request_stop()
    coord.join(threads)