python机器学习学习笔记（三）

线性回归 

1.导入线性回归模型

from sklearn import linear_model
linerg = linear_model.LinearRegression()

2.导入Diabetes数据集并分成训练集和测试集两部分

from sklearn import datasets
diabetes = datasets.load_diabetes()
x_train = diabetes.data[:-20]
y_train = diabetes.target[:-20]
x_test = diabetes.data[-20:]
y_test = diabetes.target[-20:]

3.调用分类器构造函数，用fit()函数进行训练

linreg.fit(x_train,y_train)

4.调用预测模型的coef_属性，得到每种数据的回归系数b

linreg.coef_
print(linreg.coef_)

输出b值：

[ 3.03499549e-01 -2.37639315e+02  5.10530605e+02  3.27736980e+02
 -8.14131709e+02  4.92814588e+02  1.02848452e+02  1.84606489e+02
  7.43519617e+02  7.60951722e+01]

5.在预测模型linreg上调用predict()函数进行预测并与实际值对比

linreg.predict(x_test)
print (linreg.predict(x_test))
print(y_test)

输出预测值：

[197.61846908 155.43979328 172.88665147 111.53537279 164.80054784
 131.06954875 259.12237761 100.47935157 117.0601052  124.30503555
 218.36632793  61.19831284 132.25046751 120.3332925   52.54458691
 194.03798088 102.57139702 123.56604987 211.0346317   52.60335674]

输出实际值：

[233.  91. 111. 152. 120.  67. 310.  94. 183.  66. 173.  72.  49.  64.
  48. 178. 104. 132. 220.  57.]

6.评价预测结果（越接近于1结果越准确）

linreg.score(x_test,y_test)
print(linreg.score(x_test,y_test))

0.5850753022690574

---

分析单个因素与目标值之间的回归关系并绘制图表

import numpy as np
import matplotlib.pyplot as plt
from sklearn import linear_model
from sklearn import datasets

diabetes = datasets.load_diabetes()
x_train = diabetes.data[:-20]
y_train = diabetes.target[:-20]
x_test = diabetes.data[-20:]
y_test = diabetes.target[-20:]

x0_test = x_test[:,0]
x0_train = x_train[:,0]
x0_test = x0_test[:,np.newaxis]
x0_train = x0_train[:,np.newaxis]

linreg = linear_model.LinearRegression()
linreg.fit(x0_train,y_train)
y = linreg.predict(x0_test)

plt.scatter(x0_test,y_test,color='blue')
plt.plot(x0_test,y,color='r',linewidth=3)
plt.show()

对10个因素进行回归分析并绘制图表

import numpy as np
import matplotlib.pyplot as plt
from sklearn import linear_model
from sklearn import datasets

diabetes = datasets.load_diabetes()
linreg = linear_model.LinearRegression()
x_train = diabetes.data[:-20]
y_train = diabetes.target[:-20]
x_test = diabetes.data[-20:]
y_test = diabetes.target[-20:]

plt.figure(figsize=(8,12))

for f in range(0,10):
    xi_test = x_test[:,f]
    xi_train = x_train[:,f]
    xi_test = xi_test[:,np.newaxis]  #增加维度
    xi_train = xi_train[:,np.newaxis]  #增加维度
    linreg.fit(xi_train,y_train)
    y = linreg.predict(xi_test)

    plt.subplot(5,2,f+1)
    plt.scatter(xi_test,y_test,color='b')
    plt.plot(xi_test,y,color='r',linewidth=3)
plt.show()   

---

参考:

法比奥·内利. Python数据分析实战：第2版.北京：人民邮电出版社, 2019.11.