tensorflow 实现线性回归

tensorflow 实现线性回归

首先我们构造数据

import tensorflow as tf
import numpy as np
import matplotlib.pyplot as plt
import random
n_samples =100
X = np.linspace(-3,3,n_samples)
Y = np.sin(X) + np.random.uniform(0,1,n_samples)

显示下数据：

开始构建线性模型

w = tf.Variable(tf.random_normal([1]),name='w')
b = tf.Variable(tf.random_normal([1]),name='b')


xs = tf.placeholder(tf.float32, name='X')
ys = tf.placeholder(tf.float32, name='Y')

y_pred = tf.add(tf.multiply(xs, w), b)

loss = tf.square(y_pred-ys,name='loss')

lr = 0.1
epochs = 100
optimizer = tf.train.GradientDescentOptimizer(lr).minimize(loss)


with tf.Session() as sess:
    sess.run(tf.global_variables_initializer())

    writer = tf.summary.FileWriter('./graphs/linear_reg', sess.graph)

    for epoch in range(epochs):
        total_loss = 0
        for x, y in zip(X, Y):
            _, tmp_loss= sess.run([optimizer, loss], feed_dict={xs: x, ys: y})


            total_loss += tmp_loss

        if epoch%10 == 0:

            print('Epoch {}:total_loss = {}'.format(epoch, total_loss/n_samples))

    writer.close()

    w, b = sess.run([w, b])

显示最终的结果：

运行过程：

打开tensorboard查看：