Python机器学习--回归

• 线性回归

# -*- coding: utf-8 -*-
"""
Created on Wed Aug 30 19:55:37 2017
 
@author: Administrator
"""
 
'''
背景：与房价密切相关的除了单位的房价，还有房屋的尺寸。我们可以根
据已知的房屋成交价和房屋的尺寸进行线性回归，继而可以对已知房屋尺
寸，而未知房屋成交价格的实例进行成交价格的预测
'''
 
import matplotlib.pyplot as plt
import numpy as np
from sklearn import linear_model
 
# 读取数据集
datasets_X = []
datasets_Y = []
fpath='F:\\RANJIEWEN\\MachineLearning\\Python机器学习实战_mooc\\data\\回归\\'
fr = open(fpath+'prices.txt','r')
lines = fr.readlines()
for line in lines:
    items = line.strip().split(',')
    datasets_X.append(int(items[0]))
    datasets_Y.append(int(items[1]))
 
length = len(datasets_X)
datasets_X = np.array(datasets_X).reshape([length,1])
datasets_Y = np.array(datasets_Y)
 
minX = min(datasets_X)
maxX = max(datasets_X)
X = np.arange(minX,maxX).reshape([-1,1])
 
linear = linear_model.LinearRegression()
linear.fit(datasets_X, datasets_Y)
 
# 图像中显示
plt.scatter(datasets_X, datasets_Y, color = 'red')
plt.plot(X, linear.predict(X), color = 'blue')
plt.xlabel('Area')
plt.ylabel('Price')
plt.show()

• 多项式回归

# -*- coding: utf-8 -*-
"""
Created on Wed Aug 30 20:24:09 2017
 
@author: Administrator
"""
 
'''
我们在前面已经根据已知的房屋成交价和房屋的尺寸进行了线
性回归，继而可以对已知房屋尺寸，而未知房屋成交价格的实例进行了成
交价格的预测，但是在实际的应用中这样的拟合往往不够好，因此我们在
此对该数据集进行多项式回归。
 
'''
 
import matplotlib.pyplot as plt
import numpy as np
from sklearn import linear_model
from sklearn.preprocessing import PolynomialFeatures
 
# 读取数据集
datasets_X = []
datasets_Y = []
 
fpath='F:\\RANJIEWEN\\MachineLearning\\Python机器学习实战_mooc\\data\\回归\\'
fr = open(fpath+'prices.txt','r')
lines = fr.readlines()
for line in lines:
    items = line.strip().split(',')
    datasets_X.append(int(items[0]))
    datasets_Y.append(int(items[1]))
 
length = len(datasets_X)
datasets_X = np.array(datasets_X).reshape([length,1])
datasets_Y = np.array(datasets_Y)
 
minX = min(datasets_X)
maxX = max(datasets_X)
X = np.arange(minX,maxX).reshape([-1,1])
 
poly_reg = PolynomialFeatures(degree = 2)
X_poly = poly_reg.fit_transform(datasets_X)
lin_reg_2 = linear_model.LinearRegression()
lin_reg_2.fit(X_poly, datasets_Y)
 
# 图像中显示
plt.scatter(datasets_X, datasets_Y, color = 'red')
plt.plot(X, lin_reg_2.predict(poly_reg.fit_transform(X)), color = 'blue')
plt.xlabel('Area')
plt.ylabel('Price')
plt.show()

• 岭回归

• 还有就是容易过拟合，才出现了岭回归，L2正则项

# -*- coding: utf-8 -*-
"""
Created on Wed Aug 30 20:33:00 2017
 
@author: Administrator
"""
 
'''
数据介绍：
数据为某路口的交通流量监测数据，记录全年小时级别的车流量。
实验目的：
根据已有的数据创建多项式特征，使用岭回归模型代替一般的线性模型，对
车流量的信息进行多项式回归。
'''
 
import numpy as np
 
from sklearn.linear_model import Ridge
from sklearn import cross_validation
import matplotlib.pyplot as plt
from sklearn.preprocessing import PolynomialFeatures
 
fpath='F:\RANJIEWEN\MachineLearning\Python机器学习实战_mooc\data\回归\岭回归.csv'
 
data=pd.read_csv(fpath,encoding='gbk',parse_dates=[0],index_col=0)
 
#data.sort_index(0,ascending=True,inplace=True)
 
X=data.iloc[:,:4]  ##语法
y=data.iloc[:,4]
poly=PolynomialFeatures(6) #设置多项式的最高次数
X=poly.fit_transform(X)
 
train_set_X,test_set_X,train_set_y,test_set_y=   \
    cross_validation.train_test_split(X,y,test_size=0.3,random_state=0) #设置测试集的比例，random_state随机数种子
 
clf=Ridge(alpha=1.0,fit_intercept=True)
clf.fit(train_set_X,train_set_y)
clf.score(test_set_X,test_set_y)  
 
#plot
start=200
end=300
y_pre=clf.predict(X)
time=np.arange(start,end)
plt.plot(time,y[start:end],'b',label='real')
plt.plot(time,y_pre[start:end],'r',label='predict')
plt.legend(loc='upper left')
plt.show()

• Lasso回归，添加L1正则项，具有稀疏解