Python大数据第三次的作业

---恢复内容开始---

1.逻辑回归

# 数据读取
import pandas as pd
default = pd.read_csv('Default.csv')
print(default)

## 数据预处理
## 数值编码
for item in ['student', 'default']:
    default[item] = default[item].replace({'No':0, 'Yes':1})
print(default)
## Min-Max标准化
default[['balance', 'income']] = default[['balance', 'income']].apply(lambda x:(x-x.min())/(x.max()-x.min()))
print(default)

## 训练集测试集分割
X = default.drop('default', axis=1)
y = default['default']

from sklearn.model_selection import train_test_split
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.15, random_state=10)

# 建立模型
from sklearn.linear_model import *
model_LR = LogisticRegression(class_weight='balanced', random_state=10)

# 训练模型
model_LR.fit(X_train, y_train) 

# 模型评价
model_eval = model_LR.score(X_test, y_test)
model_eval 

# 导入需要的包
import pandas as pd
from sklearn.model_selection import train_test_split
from sklearn.linear_model import LogisticRegression
from sklearn.metrics import classification_report
import seaborn as sns

# 数据处理
data = pd.read_csv('caesarian.csv')
y = data['Caesarian']
X = data.drop('Caesarian',axis=1)

# 划分训练集与测试集
x_train,x_test,y_train,y_test = train_test_split(X,y,test_size = 0.2,random_state = 10)

# 建立逻辑回归模型
clf = LogisticRegression(class_weight = 'balanced',random_state = 10,solver = 'sag')
clf.fit(x_train,y_train)
y_pred=clf.predict(x_test)

# 分类正确率
score = clf.score(x_test,y_test)
print(score)

# 分类报告
report = classification_report(y_test, y_pred)
print(report)

2.朴素贝叶斯

#######################################
# 导入需要的包
import pandas as pd
from sklearn.model_selection import train_test_split
from sklearn.naive_bayes import MultinomialNB

# 数据处理
data=pd.read_csv('caesarian.csv')
y = data['Caesarian']
X = data.drop('Caesarian',axis = 1) 

# 划分训练集与测试集
x_train,x_test,y_train,y_test = train_test_split(X,y,test_size=0.2,random_state=10)

# 建立模型
MNB = MultinomialNB(alpha = 0.01)
clf = MNB.fit(x_train, y_train)

# 输出准确率
score = clf.score(x_test,y_test)
print(score)


#######################################
# 导入需要的包
import pandas as pd
import numpy as np

# 数据处理
data = pd.read_csv('caesarian.csv')
y = data['Caesarian']
X = data[['Age','Delivery Number','Delivery time','Blood of Pressure','Heart Problem']]

# 划分训练集与测试集
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,random_state = 10)

# 按照线性核、多项式核、Sigmoid核、高斯核的顺序，分别选用不同的核函数
from sklearn.svm import SVC

kernels=['linear','poly','sigmoid','rbf']
kernel_scores = []

for kernel in kernels:
    kernel_score = SVC(kernel = kernel,random_state=10).fit(x_train,y_train).score(x_test,y_test)
    kernel_scores.append(kernel_score)
    
print(kernel_scores)

3.支持向量机

# 导入需要的包
import pandas as pd
import numpy as np

# 数据处理
data=pd.read_csv('caesarian.csv')
caesarian=data['Caesarian']
feature=['Age','Delivery Number','Delivery time','Blood of Pressure','Heart Problem']
traindata=data[feature]
print(traindata.head(4))

# 划分训练集与测试集
from sklearn.model_selection import train_test_split
x_train,x_test,y_train,y_test=train_test_split(traindata,caesarian,train_size=0.8,random_state=10)

# 载入模型
from sklearn.svm import LinearSVC

# 建立模型
lsvm=LinearSVC(C=0.68,random_state=10)

# 训练模型
lsvm.fit(x_train,y_train)

# 模型评价
score=lsvm.score(x_test,y_test)
print(score)

# 导入需要的包
import pandas as pd
from sklearn.model_selection import train_test_split
from sklearn.svm import SVC

# 数据处理
data = pd.read_csv('caesarian.csv')
y = data['Caesarian']
X = data.drop('Caesarian',axis = 1)

# 划分训练集与测试集
x_train,x_test,y_train,y_test = train_test_split(X,y,test_size=0.2,random_state = 10)

# 调整支持向量机模型的参数
clf = SVC(kernel='linear',verbose = True)
clf.fit(x_train,y_train)

# 分类正确率
score = clf.score(x_test,y_test)
print(score)

# 支持向量
SV = clf.support_vectors_
print('\n\n',SV)

# 正类和负类支持向量的索引
S = clf.support_ 
print('\n\n',S)

# 每个类支持向量的个数
NS = clf.n_support_
print('\n\n',NS)

# 超平面系数
C = clf.coef_
print('\n\n',C)

 

4.K近邻

# 导入需要的包
import pandas as pd
import numpy as np

# 数据处理
data=pd.read_csv('caesarian.csv')
caesarian=data['Caesarian']
feature=['Age','Delivery Number','Delivery time','Blood of Pressure','Heart Problem']
traindata=data[feature]

# 划分训练集与测试集
from sklearn.model_selection import train_test_split
x_train,x_test,y_train,y_test=train_test_split(traindata,caesarian,train_size=0.8,random_state=10)

# 导入模型
from sklearn.neighbors import *

# 输出分类正确率
K_scores = [KNeighborsClassifier(n_neighbors=k).fit(x_train,y_train).score(x_test,y_test) for k in range(2,10,2)]
print(K_scores)

5.决策树

#  数据处理
data=pd.read_csv('caesarian.csv')
caesarian=data['Caesarian']
feature=['Age','Delivery Number','Delivery time','Blood of Pressure','Heart Problem']
traindata=data[feature]

# 划分训练集与测试集
from sklearn.model_selection import *
x_train,x_test,y_train,y_test=train_test_split(traindata, caesarian, test_size=0.2, random_state=10)

# 训练模型
from sklearn.tree import *
DF_model = DecisionTreeClassifier(random_state=10)
DF_model.fit(x_train, y_train)

# 模型预测
result = DF_model.predict(x_test)
result = round(DF_model.score(x_test, y_test),4)
print(result)
# 查看各特征的重要性并绘图
subplot=pd.Series(DF_model.feature_importances_, index=x_train.columns).sort_values().plot(kind='barh',  title='特征重要性')

 

---恢复内容结束---