scikit-learn的线性回归+岭回归+Logistics回归

普通的最小二乘法

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

#生成列向量
x0 = np.random.normal(size=109).T
x1 = np.random.normal(size=109).T
x2 = np.random.normal(size=109).T

y = 2*x1 + 3*x2 + x0*9
#合并矩阵
X = np.column_stack((x0, x1,x2))

#拟合
reg = linear_model.LinearRegression()
reg.fit (X[:50],y[:50])

y_pred = reg.predict(X[50:])
print(reg.coef_)

岭回归

如果不用岭回归，效果如下

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

x0 = np.random.normal(size=109).T
x1 = np.random.normal(size=109).T

x2 = x0 + x1


y = 2*x1 + 3*x2 + x0*6
#合并列向量
X = np.column_stack((x0, x1,x2))


reg = linear_model.LinearRegression()
reg.fit (X[:80],y[:80])

print(reg.coef_)

此时输出结果为[4.33333333 0.33333333 4.66666667]，不符合预期
采用岭回归

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

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

x0 = np.random.normal(size=109).T
x1 = np.random.normal(size=109).T
x2 = x0 + x1


y = 2*x1 + 3*x2 + x0*9
X = np.column_stack((x0, x1,x2))

reg = reg = linear_model.RidgeCV(alphas=[0.0001, 1.0, 100.0])
reg.fit (X[:80],y[:80])

print(reg.coef_)
print(reg.intercept_)

此时输出结果为[ 6.33332712 -0.66666281 5.66666431]，其实也没多符合。

Logistics回归

from sklearn.linear_model import LogisticRegression
from sklearn import metrics
import numpy as np 
from matplotlib import pyplot as plt 

X0 = [
    [-1,1],
    [-1,2],
    [-2,3],
    [-0.5,6],
    [-5,2],
    [-10,1],
    [-19,2,],
    [-5,-1],
    [-6,-5],
    [-7,-8],
    [-2,-11],
    [2,2],
    [3,9],
    [4,1],
    [5,1],
    [6,5],
    [4,3],
    [2,6],
    [3,-1],
    [6,-2],
    [6,-9],
    [1,-2],
    [1,-1],
    [2,-3],
    [3,-6],
    [7,-5],
]

X = np.zeros(26)
Y = np.zeros(26)
for i in range(26):
    X[i]=X0[i][0]
    Y[i]=X0[i][1]

y = [
    0,0,0,0,0,0,0,
    1,1,1,1,
    2,2,2,2,2,2,2,
    3,3,3,3,3,3,3,3
]



reg = LogisticRegression()
reg.fit(X0,y)

X_test = [[1,1],[-1,-1],[1,-1],[-1,1]]
print(reg.predict(X_test))


plt.figure(1)
ax = plt.axes()
ax.set_xlim(-10,10)
ax.set_ylim(-10,10)
plt.scatter(X[:7],Y[:7])
plt.scatter(X[7:11],Y[7:11])
plt.scatter(X[11:18],Y[11:18])
plt.scatter(X[18:26],Y[18:26])

plt.show()