使用scikit-learn构建模型
sklearn中还存在许多不同的机器学习模型可以直接调用，相比于自己撰写代码，直接使用sklearn的模型可以大大提高效率。
sklearn中所有的模型都有四个固定且常用的方法，分别是model.fit、model.predict、model.get_params、model.score。
1.数据导入
from sklearn.datasets import load_boston
from sklearn.model_selection import train_test_split
from sklearn.preprocessing import MinMaxScaler
from sklearn.decomposition import PCA
data=load_boston()
2.数据预处理
data.keys()
Out[ ]:
dict_keys(['data', 'target', 'feature_names', 'DESCR', 'filename', 'data_module'])
In [ ]:
data['data']
Out[ ]:
array([[6.3200e-03, 1.8000e+01, 2.3100e+00, ..., 1.5300e+01, 3.9690e+02,
4.9800e+00],
[2.7310e-02, 0.0000e+00, 7.0700e+00, ..., 1.7800e+01, 3.9690e+02,
9.1400e+00],
[2.7290e-02, 0.0000e+00, 7.0700e+00, ..., 1.7800e+01, 3.9283e+02,
4.0300e+00],
...,
[6.0760e-02, 0.0000e+00, 1.1930e+01, ..., 2.1000e+01, 3.9690e+02,
5.6400e+00],
[1.0959e-01, 0.0000e+00, 1.1930e+01, ..., 2.1000e+01, 3.9345e+02,
6.4800e+00],
[4.7410e-02, 0.0000e+00, 1.1930e+01, ..., 2.1000e+01, 3.9690e+02,
7.8800e+00]])
In [ ]:
#将data['data']和data['target']两类数据都切割为两份，test_size=0.2
train_data,test_data,train_target,test_target=train_test_split(data['data'],data['target'],test_size=0.2)
print(train_data.shape)
print(test_data.shape)
print(train_target.shape)
print(test_target.shape)
(404, 13)
(102, 13)
(404,)
(102,)
3.构建离差标准化模型
#构建离差标准化模型
model=MinMaxScaler().fit(train_data)
In [ ]:
model.transform(train_data)
Out[ ]:
array([[6.17624751e-04, 4.00000000e-01, 2.18108504e-01, ...,
5.31914894e-01, 9.91252092e-01, 6.70529801e-02],
[9.15525569e-02, 0.00000000e+00, 6.46627566e-01, ...,
8.08510638e-01, 1.00000000e+00, 5.27593819e-01],
[1.85321145e-03, 0.00000000e+00, 2.96920821e-01, ...,
8.82978723e-01, 9.96881181e-01, 4.67163355e-01],
...,
[2.70765792e-04, 0.00000000e+00, 1.73387097e-01, ...,
8.08510638e-01, 9.94700543e-01, 2.43653422e-01],
[6.04923824e-03, 0.00000000e+00, 7.85557185e-01, ...,
9.14893617e-01, 1.00000000e+00, 4.61644592e-01],
[3.85613648e-03, 0.00000000e+00, 3.46041056e-01, ...,
6.17021277e-01, 9.98326487e-01, 2.27373068e-01]])
In [ ]:
train_data_mms=model.transform(train_data)
In [ ]:
model.transform(test_data)
Out[ ]:
array([[1.10386796e-02, 0.00000000e+00, 2.81524927e-01, ...,
8.93617021e-01, 9.94015924e-01, 5.00827815e-01],
[4.82624007e-03, 0.00000000e+00, 3.71334311e-01, ...,
6.38297872e-01, 1.00000000e+00, 5.89403974e-01],
[2.76127157e-03, 0.00000000e+00, 3.71334311e-01, ...,
6.38297872e-01, 9.81058877e-01, 4.50607064e-01],
...,
[3.27751369e-04, 8.00000000e-01, 1.16568915e-01, ...,
4.04255319e-01, 9.95638724e-01, 2.07505519e-01],
[8.61718595e-03, 0.00000000e+00, 2.81524927e-01, ...,
8.93617021e-01, 9.77280795e-01, 3.05463576e-01],
[2.65516824e-02, 0.00000000e+00, 7.00879765e-01, ...,
2.23404255e-01, 9.86840103e-01, 7.67108168e-01]])
In [ ]:
test_data_mms=model.transform(test_data)
In [ ]:
#利用PCA进行降维
model=PCA(n_components=8).fit(train_data_mms)
In [ ]:
model.transform(train_data_mms).shape
Out[ ]:
(404, 8)
In [ ]:
train_data_mms=model.transform(train_data_mms)
In [ ]:
model.transform(test_data_mms).shape
Out[ ]:
(102, 8)
In [ ]:
test_data_mms=model.transform(test_data_mms)
4.聚类分析
#使用sklearn估计器进行聚类分析
from sklearn.datasets import load_iris
from sklearn.cluster import KMeans
data=load_iris()
#使用KMeans聚类模型将数据data['data']数据分为三类
KMeans(n_clusters=3).fit(data['data'])
Out[ ]:
KMeans
KMeans(n_clusters=3)
In [ ]:
model=KMeans(n_clusters=3).fit(data['data'])
#可以查看聚类标签
model.labels_
Out[ ]:
array([1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
1, 1, 1, 1, 1, 1, 2, 2, 0, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2,
2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 0, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2,
2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 0, 2, 0, 0, 0, 0, 2, 0, 0, 0,
0, 0, 0, 2, 2, 0, 0, 0, 0, 2, 0, 2, 0, 2, 0, 0, 2, 2, 0, 0, 0, 0,
0, 2, 0, 0, 0, 0, 2, 0, 0, 0, 2, 0, 0, 0, 2, 0, 0, 2])
In [ ]:
#可以查看聚类中心
model.cluster_centers_
Out[ ]:
array([[6.85 , 3.07368421, 5.74210526, 2.07105263],
[5.006 , 3.428 , 1.462 , 0.246 ],
[5.9016129 , 2.7483871 , 4.39354839, 1.43387097]])
In [ ]:
#通过可视化查看聚类效果
import matplotlib.pyplot as plt
for i in range(3):
plt.scatter(data['data'][model.labels_==i,0],data['data'][model.labels_==i,1])
plt.show()
In [ ]:
#聚类模型评价
from sklearn.metrics import silhouette_score
silhouette_score(data['data'],model.labels_)
#分析聚为几类比较好
for k in range(2,9):
model=KMeans(n_clusters=k).fit(data['data'])
print(k,silhouette_score(data['data'],model.labels_))
2 0.6810461692117462
3 0.5528190123564095
4 0.49805050499728737
5 0.48874888709310566
6 0.36483400396700255
7 0.348950840496829
8 0.3617900335973811
5.分类模型
#使用sklearn估计器进行分类分析
#导入数据
from sklearn.datasets import load_breast_cancer
data=load_breast_cancer()
#获取数据
x=data['data']
#是否是乳腺癌
y=data['target']
In [ ]:
#划分训练集，测试集
from sklearn.model_selection import train_test_split
x_train,x_test,y_train,y_test=train_test_split(x,y,test_size=0.2)
#模型预处理
x_train.shape
Out[ ]:
(455, 30)
In [ ]:
#查看每一列的最大值
import numpy as np
#将科学计数法转为数值
np.int32(x_train.max(axis=0))
Out[ ]:
array([ 28, 39, 188, 2501, 0, 0, 0, 0, 0, 0, 2,
4, 21, 542, 0, 0, 0, 0, 0, 0, 36, 49,
251, 4254, 0, 0, 1, 0, 0, 0])
In [ ]:
#对数据进行标准化离散标准化处理
from sklearn.preprocessing import StandardScaler
model=StandardScaler().fit(x_train)
x_train_ss=model.transform(x_train)
x_test_ss=model.transform(x_test)
In [ ]:
#检验标准化后的最大值结果
x_train_ss.max(axis=0)
#分类模型构建
from sklearn.svm import SVC
#构建SVC模型
model=SVC().fit(x_train_ss,y_train)
#对模型进行预测，并输出预测结果
model.predict(x_test_ss)
Out[ ]:
array([1, 0, 0, 1, 1, 0, 1, 1, 1, 0, 0, 1, 1, 1, 0, 1, 1, 1, 1, 0, 1, 1,
1, 0, 1, 0, 1, 1, 1, 0, 0, 1, 1, 0, 0, 1, 1, 1, 1, 1, 0, 1, 1, 0,
0, 0, 1, 0, 1, 0, 0, 1, 1, 1, 1, 1, 0, 0, 0, 1, 0, 0, 1, 1, 1, 0,
1, 1, 1, 0, 0, 0, 1, 1, 0, 0, 0, 0, 1, 0, 0, 1, 1, 1, 0, 1, 0, 1,
1, 0, 1, 0, 0, 1, 0, 0, 0, 1, 1, 1, 0, 1, 1, 0, 1, 1, 0, 1, 0, 1,
1, 0, 1, 0])
In [ ]:
#再看和真实值对比，看泛化能力
y_test
Out[ ]:
array([1, 0, 0, 1, 1, 0, 1, 1, 1, 0, 0, 1, 1, 1, 0, 1, 1, 1, 1, 0, 1, 1,
1, 0, 1, 0, 1, 1, 1, 0, 0, 1, 1, 0, 0, 1, 1, 1, 1, 1, 0, 1, 1, 0,
0, 0, 1, 0, 1, 0, 0, 1, 1, 1, 1, 1, 0, 0, 0, 1, 0, 0, 1, 1, 1, 0,
1, 1, 1, 0, 0, 0, 1, 1, 0, 0, 0, 0, 1, 0, 0, 1, 1, 1, 0, 1, 0, 1,
1, 0, 1, 0, 0, 1, 0, 0, 0, 1, 1, 1, 0, 1, 1, 0, 1, 1, 0, 1, 0, 1,
1, 0, 1, 0])
In [ ]:
#直接计算真实值与训练值的准确率
y_pre=model.predict(x_test_ss)
model.score(x_test_ss,y_test)
Out[ ]:
1.0
In [ ]:
#分类模型的评价指标
from sklearn.metrics import recall_score,precision_score,f1_score,roc_curve
#召回率
print(recall_score(y_test,y_pre))
#准确率
print(precision_score(y_test,y_pre))
#综合衡量召回率和准确率
print(f1_score(y_test,y_pre))
1.0
1.0
1.0
In [ ]:
fpr,tpr,thresholds=roc_curve(y_test,y_pre)
import matplotlib.pyplot as plt
plt.plot(fpr,tpr)
plt.show()
6.回归模型
#使用sklearn估计器构建回归模型
from sklearn.datasets import load_boston
data=load_boston()
x=data['data']
y=data['target']
from sklearn.model_selection import train_test_split
x_train,x_test,y_train,y_test=train_test_split(x,y,test_size=0.2)
In [ ]:
from sklearn.linear_model import LinearRegression
model=LinearRegression().fit(x_train,y_train)
In [ ]:
model.predict(x_test)
Out[ ]:
array([25.63512704, 18.20664391, 19.30419437, 17.56173266, 32.65076406,
20.51863333, 21.18307155, 17.69644861, 7.17364282, 21.24312946,
9.31181655, 22.58174485, 24.86143471, 17.14942807, 18.75445808,
32.02048648, 3.20366619, 18.45078084, 24.47301112, 19.91054657,
16.04118568, 23.7397113 , 7.84796407, 15.62557098, 27.29096309,
30.82585562, 22.4279634 , 20.70777259, 18.12353173, 18.57206891,
13.03685438, 12.61133503, 22.83018416, 16.01519429, 16.82163751,
24.51813482, 15.78043774, 23.40462434, 17.11255973, 28.93962029,
22.72251851, 33.18948449, 20.98919608, 20.07592808, 34.36482708,
25.23664217, 21.55366314, 15.94069443, 25.59419884, 14.1418572 ,
23.33774419, 25.04689547, 23.29641008, 33.950857 , 25.18734214,
7.63796202, 29.91317932, 13.18689018, 25.01891632, 31.05545963,
17.0977797 , 29.67337528, 36.4307191 , 16.05518381, 17.54582224,
32.33404942, 32.77216924, 18.44714854, 16.77438163, 17.65500137,
8.95855944, 27.58690131, 12.20201901, 24.89726481, 30.49544591,
21.85276613, 23.00784191, 31.17132898, 16.77202264, 18.96818273,
16.02055635, 24.45445282, 15.616554 , 22.85427056, 20.25649892,
18.5264222 , 21.18237716, 27.5588005 , 23.86810489, 16.56091536,
16.47704557, 26.20295992, 27.99514403, 10.97460741, 5.88779974,
32.77537677, 32.03321212, 15.50207643, 32.37743616, 21.86224154,
13.03662501, 21.32844856])
In [ ]:
y_pre=model.predict(x_test)
In [ ]:
import matplotlib.pyplot as plt
plt.plot(range(len(y_test)),y_test)
plt.plot(range(len(y_pre)),y_pre)
plt.legend(['real','predict'])
plt.show()
In [ ]:
from sklearn.metrics import mean_squared_error,r2_score
mean_squared_error(y_true=y_test,y_pred=y_pre)
r2_score(y_true=y_test,y_pred=y_pre)
Out[ ]:
0.5489931237687122
7.参考文章
机器学习好伙伴之scikit-learn的使用——常用模型及其方法
【创作不易，望点赞收藏，若有疑问，请留言，谢谢】
