0
点赞
收藏
分享

微信扫一扫

利用Tensorflow来实现线性回归算法的简单案例

崭新的韭菜 2022-01-13 阅读 51
import numpy as np
x_data=np.float32(np.random.rand(100,2))
y_data=(np.dot(x_data,[0.100,0.200])+0.300).reshape(-1,1)

import tensorflow as tf
x_input=tf.placeholder(tf.float32,shape=[None,2])
y_actual=tf.placeholder(tf.float32,shape=[None,1])
W=tf.Variable(tf.random_uniform([2,1],-1.0,1.0))
b=tf.Variable(tf.zeros([1]))
m=tf.matmul(x_input,W)
y=m+b

loss=tf.reduce_mean(tf.square(y-y_actual))
optimizer=tf.train.GradientDescentOptimizer(0.5)
train=optimizer.minimize(loss)

init=tf.global_variables_initializer()
sess=tf.Session()
sess.run(init)

for step in range(0,200):
    sess.run(train,feed_dict={x_input:x_data,y_actual:y_data})
    if step%20==0:
        print(step,sess.run(W).T[0],sess.run(b))

1.生成数组(line1-3)

2.构建模型(line5-11)

3.确定损失函数(line13)

4.设置学习率(line14)

5.设定模型训练目标(line15)

6.模型初始化(line17-19)

7.模型训练(line21-24)

举报

相关推荐

0 条评论