Tensorflow学习笔记：线性回归简单实现

#coding:utf-8
'''
线性回归
'''

import numpy as np
import matplotlib.pyplot as plt
import tensorflow as tf

# 定义线性回归模型
def model(X,w):
    return tf.multiply(X,w)

x_train = np.linspace(-1,1,101)
y_train = 2 * x_train + np.random.randn(*x_train.shape) * 0.33

learning_rate = 0.01
training_epochs = 100

X = tf.placeholder(tf.float32)
Y = tf.placeholder(tf.float32)

w = tf.Variable(0.0,name='weights')

y_model = model(X,w)

# 定义损失函数
cost = tf.square(Y - y_model)
# 使用梯度下降法训练
train_op = tf.train.GradientDescentOptimizer(learning_rate).minimize(cost) 
init_op = tf.global_variables_initializer()

sess = tf.Session()

sess.run(init_op)

# 训练
for epoch in range(training_epochs):
    for (x,y) in zip(x_train,y_train):
        sess.run(train_op,feed_dict={X:x,Y:y})

w_val = sess.run(w)
sess.close()
print('value of w:',w_val)

# 显示拟合结果
plt.scatter(x_train, y_train) 
y_learned = x_train*w_val 
plt.plot(x_train, y_learned, 'r') 
plt.show()