学习进度笔记

学习进度笔记13

TensorFlow编程基础——实现线性回归

# 载入必要库

import tensorflow as tf

import numpy as np

import matplotlib.pyplot as plt

# 设置必要参数

 

## 设置学习率

learning_rate = 0.01

 

## 设置迭代轮数

training_epochs = 1000

 

## 每50轮展示当前模型的参数值和损失

display_step = 50

 

## 每500轮保存一次模型

save_step = 500

# 设定原始数据

 

## 训练集

train_X = np.asarray([3.3,4.4,5.5,6.71,6.93,4.168,9.779,6.182,7.59,2.167,

                         7.042,10.791,5.313,7.997,5.654,9.27,3.1])

 

## 训练集标签

train_Y = np.asarray([1.7,2.76,2.09,3.19,1.694,1.573,3.366,2.596,2.53,1.221,

                         2.827,3.465,1.65,2.904,2.42,2.94,1.3])

# 定义张量占位符

X = tf.placeholder("float", name="X")

Y = tf.placeholder("float", name="Y")

# 定义权重和偏置

 

with tf.variable_scope("liner_regression"):

    # 设置模型的权重和偏置

    W = tf.get_variable(initializer=np.random.randn(), name="weight")  # 生成权重

    b = tf.get_variable(initializer=np.random.randn(), name="bias")  # 生成偏置

    # 构建线性回归模型（前向传播）

    mul = tf.multiply(X, W, name="mul")

pred = tf.add(mul, b, name="pred")

# 建立会话运行程序

with tf.Session() as sess:

 

    # 初始化变量

    init_op = tf.global_variables_initializer()

    sess.run(init_op)

    

    # 将汇总结果写入文件

    file_writer = tf.summary.FileWriter("./temp/summary/linear", graph=sess.graph)

 

    # 拟合训练数据

    for epoch in range(training_epochs):

        for (x, y) in zip(train_X, train_Y):

            # 带入数据

            _, summary = sess.run([train_op, merged], feed_dict={X: x, Y: y})

 

        # 保存模型

        if (epoch + 1) % save_step == 0:

            save_path = saver.save(sess, ckpt_path, global_step=epoch)

            print("Model saved in file: %s" % save_path)

 

        # 展示每步训练的日志

        if (epoch + 1) % display_step == 0:

            # Display loss and value

            c = sess.run(loss, feed_dict={X: train_X, Y: train_Y})

            print("Epoch:", '%04d' % (epoch+1), "loss=", "{:.9f}".format(c), "W=", W.eval(), "b=", b.eval())

 

            file_writer.add_summary(summary, global_step=epoch)

 

    print("Optimization Finished!")

 

    # 保存最终模型

    save_path = saver.save(sess, ckpt_path, global_step=epoch)

    print("Final model saved in %s" % save_path)

 

    # 计算最终损失函数

    training_loss = sess.run(loss, feed_dict={X: train_X, Y: train_Y})

    print("Training loss=", training_loss, "W=", sess.run(W), "b=", sess.run(b), '\n')

 

    # 画图

    plt.plot(train_X, train_Y, 'ro', label='Original data')

    plt.plot(train_X, sess.run(W) * train_X + sess.run(b), label='Fitted line')

    plt.legend()

    plt.show()