TensorFlow基础3-机器学习基础知识（解析法实现一元线性回归、多元线性回归）

记录TensorFlow听课笔记

---

一，机器学习基础

1.1一元线性回归

模型：y =wx+b
模型变量：x 模型参数
w：权重（weights）
b：偏置值（bias）

1.2解析法实现一元线性回归

1.3解析法实现多元线性回归

二，代码实现一元/多元回归

2.1解析法实现一元线性回归代码实现

导入库，设置字体，加载样本数据（numpy方法实现）

import numpy as np
x=np.array([137.97,104.50,100.00,124.32,79.20,99.00,124.00,114.00,106.69,138.05,53.75,46.91,68.00,63.02,81.26,86.21])
y=np.array([145.00,110.00,93.00,116.00,65.32,104.00,118.00,91.00,62.00,133.00,51.00,45.00,78.50,69.65,75.69,95.30])
meanX=np.mean(x)
meanY=np.mean(y)
sumXY=np.sum((x-meanX)*(y-meanY))
sumX=np.sum((x-meanX)*(x-meanX))
w=sumXY/sumX
b=meanY-w*meanX
print("w=",w)
print("b=",b)
x_test=np.array([128.15,45.00,141.43,106.27,99.00,53.84,85.36,70.00])
y_pred = w*x_test+b
print("面积\t估计房价")
for i in range(y_pred.size):
    print(x_test[i],"\t",np.round(y_pred[i],2))

导入库，设置字体，加载样本数据（tf方法实现）

import tensorflow as tf
x=tf.constant([137.97,104.50,100.00,124.32,79.20,99.00,124.00,114.00,106.69,138.05,53.75,46.91,68.00,63.02,81.26,86.21])
y=tf.constant([145.00,110.00,93.00,116.00,65.32,104.00,118.00,91.00,62.00,133.00,51.00,45.00,78.50,69.65,75.69,95.30])
meanX=tf.reduce_mean(x)
meanY=tf.reduce_mean(y)
sumXY=tf.reduce_sum((x-meanX)*(y-meanY))
sumX=tf.reduce_sum((x-meanX)*(x-meanX))
w=sumXY/sumX
b=meanY-w*meanX
print("w=",w)
print("b=",b)
x_test=tf.constant([128.15,45.00,141.43,106.27,99.00,53.84,85.36,70.00])
y_pred = w*x_test+b
print(y_pred)

解析法实现一元线性回归

#解析法实现一元线性回归 
#导入库
import tensorflow as tf
import numpy as np
import matplotlib.pyplot as plt 
#设置字体
plt.rcParams['font.sans-serif'] =['SimHei']
#加载样本数据
x=tf.constant([137.97,104.50,100.00,124.32,79.20,99.00,124.00,114.00,106.69,138.05,53.75,46.91,68.00,63.02,81.26,86.21])
y=tf.constant([145.00,110.00,93.00,116.00,65.32,104.00,118.00,91.00,62.00,133.00,51.00,45.00,78.50,69.65,75.69,95.30])
#学习模型-计算w,b
meanX=tf.reduce_mean(x)
meanY=tf.reduce_mean(y)
sumXY=tf.reduce_sum((x-meanX)*(y-meanY))
sumX=tf.reduce_sum((x-meanX)*(x-meanX))
w=sumXY/sumX
b=meanY-w*meanX
print("权值w=",w.numpy())
print("偏置b=",b.numpy())
print("线性模型：y=",w.numpy(),"*x+",b.numpy())
#预测房价
x_test=np.array([128.15,45.00,141.43,106.27,99.00,53.84,85.36,70.00])
y_pred = (w*x_test+b).numpy()#将张量类型转化为numpy数组类型
print("面积\t估计房价")
n=len(x_test)
for i in range(n):
    print(x_test[i],"\t",round(y_pred[i],2))
#模型和数据可视化
plt.figure()
#绘制散点图
#张量和数组都可以作为散点函数的输入提供点坐标
plt.scatter(x,y,color="red",label="销售记录")
plt.scatter(x_test,y_pred,color="blue",label="预测房价")
plt.plot(x_test,y_pred,color="green",label="拟合直线",linewidth=2)
#设置坐标轴的标签文字和字号
plt.xlabel("面积（平方米）",fontsize=14)
plt.xlabel("价格（万元）",fontsize=14)
#设置坐标轴的范围
plt.xlim((40,150))
plt.ylim((40,150))
#设置标题文字和字号
plt.suptitle("商品房销售价格评估系统v1.0",fontsize=20)
#在左上方显示图例
plt.legend(loc="upper left")
#显示整个绘图
plt.show()

2.2解析法实现多元线性回归代码实现（输入平米和房间数返回预测房价）

#解析法实现多元线性回归
#导入库
import numpy as np
#=======================【1】加载样本数据===============================================
x1=np.array([137.97,104.50,100.00,124.32,79.20,99.00,124.00,114.00,106.69,138.05,53.75,46.91,68.00,63.02,81.26,86.21])
x2=np.array([3,2,2,3,1,2,3,2,2,3,1,1,1,1,2,2])
y=np.array([145.00,110.00,93.00,116.00,65.32,104.00,118.00,91.00,62.00,133.00,51.00,45.00,78.50,69.65,75.69,95.30])
#=======================【2】数据处理===============================================
x0=np.ones(len(x1))
#堆叠属性数组，构造属性矩阵
#从(16,)到(16,3),因为新出现的轴是第二个轴所以axis为1
X=np.stack((x0,x1,x2),axis=1)
print(X)
#得到形状为一列的数组
Y=np.array(y).reshape(-1,1)
print(Y)
#=======================【3】求解模型参数===============================================
Xt=np.transpose(X)                      #计算X'
XtX_1=np.linalg.inv(np.matmul(Xt,X))    #计算(X'X)-1
XtX_1_Xt=np.matmul(XtX_1,Xt)            #计算(X'X)-1X'
W=np.matmul(XtX_1_Xt,Y)                 #W=((X'X)-1)X'Y
print(W)
W=W.reshape(-1)
print(W)
print("多元线性回归方程")
print("Y=",W[1],"*x1+",W[2],"*x2+",W[0])
#=======================【4】预测房价===============================================
print("请输入房屋面积和房间数，预测房屋销售价格")
x1_test=float(input("商品房面积:"))
x2_test=int(input("房间数:"))
y_pred=W[1]*x1_test+W[2]*x2_test+W[0]
print("预测价格：",round(y_pred,2),"万元")