通过５折交叉验证，实现逻辑回归，决策树，SVM,随机森林，GBDT,Xgboost,lightGBM的评分

通过５折交叉验证，实现逻辑回归，决策树，SVM,随机森林，GBDT,Xgboost,lightGBM的评分

• 导入的包

import pandas as pd
import warnings
from sklearn.preprocessing import scale
from sklearn.model_selection import cross_val_score
from sklearn.linear_model import LogisticRegression
from sklearn.tree import DecisionTreeClassifier
from sklearn.svm import SVC
from sklearn.ensemble import RandomForestClassifier
from sklearn.ensemble import GradientBoostingClassifier
from xgboost.sklearn import XGBClassifier
import lightgbm as lgb

• 读取数据集

data_all = pd.read_csv('/home/infisa/wjht/project/DataWhale/data_all.csv', encoding='gbk')

• 处理数据集

df_y=data_all['status']
df_X=data_all.drop(columns=['status'])
df_X=scale(df_X,axis=0)  #将数据转化为标准数据

• 构建模型

lr = LogisticRegression(random_state=2018,tol=1e-6)  # 逻辑回归模型

tree = DecisionTreeClassifier(random_state=2018) #决策树模型

svm = SVC(probability=True,random_state=2018,tol=1e-6)  # SVM模型

forest=RandomForestClassifier(n_estimators=100,random_state=2018) #　随机森林

Gbdt=GradientBoostingClassifier(random_state=2018) #CBDT

Xgbc=XGBClassifier(random_state=2018)  #Xgbc

gbm=lgb.LGBMClassifier(random_state=2018)  #lgb

• 构建评分函数，并采取５折交叉验证的方式评分

def muti_score(model):
    warnings.filterwarnings('ignore')
    accuracy = cross_val_score(model, df_X, df_y, scoring='accuracy', cv=5)
    precision = cross_val_score(model, df_X, df_y, scoring='precision', cv=5)
    recall = cross_val_score(model, df_X, df_y, scoring='recall', cv=5)
    f1_score = cross_val_score(model, df_X, df_y, scoring='f1', cv=5)
    auc = cross_val_score(model, df_X, df_y, scoring='roc_auc', cv=5)
    print("准确率:",accuracy.mean())
    print("精确率:",precision.mean())
    print("召回率:",recall.mean())
    print("F1_score:",f1_score.mean())
    print("AUC:",auc.mean())

其中mean()指的是求得的均值

模型	准确率	精确率	召回率	F1_score	AUC
逻辑回归	0.7890191148682617	0.6542724662896913	0.3377975457965613	0.44525012166067884	0.7840451024530857
决策树	0.6962524533638791	0.6962524533638791	0.6962524533638791	0.6962524533638791	0.6962524533638791
SVM	0.787758390223099	0.7351623295760905	0.24060335431243626	0.36179547264664874	0.7640376541388867
随机森林	0.7921756804332226	0.7135700690071172	0.2867128441334693	0.40835414886475174	0.7752164698827589
GBDT	0.7938590063951863	0.6604108594441386	0.36633732991104395	0.4708811551285791	0.7888240065764295
Xgboost	0.7982740847293591	0.6829783239831001	0.3663162336064133	0.47673826685376613	0.7914190511145234
LightGbm	0.79049080811139	0.6421783397519263	0.3730354066312717	0.47150438344663004	0.7776116341798183

• 分析
  模型的评分思想，是通过采用５折交叉验证，得出其中的均值分数来评判。从上表中可以看出逻辑回归，随机森林，GBDT，Xgboost,LightGbm的各个指标都很相近而且分数也较高，说明这几个模型拟合数据效果都较好，都可以选做模型。综合来看Xgboost的分数更高一些，它的评分效果最好。

• 问题
  01 还没有学会在代码中，可以直接输出表格的操作。
  02 对各个模型的参数还不太了解
  03 对数据集划分，怎样构造优质的数据还比较欠缺。

• 参考文章
  cross_val_score的 scoring参数值解析
  python机器学习库sklearn——交叉验证（K折、留一、留p、随机）
  12号同学写的博客

• 完整代码

import pandas as pd
import warnings
from sklearn.preprocessing import scale
from sklearn.model_selection import cross_val_score
from sklearn.linear_model import LogisticRegression
from sklearn.tree import DecisionTreeClassifier
from sklearn.svm import SVC
from sklearn.ensemble import RandomForestClassifier
from sklearn.ensemble import GradientBoostingClassifier
from xgboost.sklearn import XGBClassifier
import lightgbm as lgb


# 读取数据集
data_all = pd.read_csv('/home/infisa/wjht/project/DataWhale/data_all.csv', encoding='gbk')

# 划分为5折交叉验证数据集
df_y=data_all['status']
df_X=data_all.drop(columns=['status'])
df_X=scale(df_X,axis=0)  #将数据转化为标准数据
#构建模型

lr = LogisticRegression(random_state=2018,tol=1e-6)  # 逻辑回归模型

tree = DecisionTreeClassifier(random_state=2018) #决策树模型

svm = SVC(probability=True,random_state=2018,tol=1e-6)  # SVM模型

forest=RandomForestClassifier(n_estimators=100,random_state=2018) #　随机森林

Gbdt=GradientBoostingClassifier(random_state=2018) #CBDT

Xgbc=XGBClassifier(random_state=2018)  #Xgbc

gbm=lgb.LGBMClassifier(random_state=2018)  #lgb



def muti_score(model):
    warnings.filterwarnings('ignore')
    accuracy = cross_val_score(model, df_X, df_y, scoring='accuracy', cv=5)
    precision = cross_val_score(model, df_X, df_y, scoring='precision', cv=5)
    recall = cross_val_score(model, df_X, df_y, scoring='recall', cv=5)
    f1_score = cross_val_score(model, df_X, df_y, scoring='f1', cv=5)
    auc = cross_val_score(model, df_X, df_y, scoring='roc_auc', cv=5)
    print("准确率:",accuracy.mean())
    print("精确率:",precision.mean())
    print("召回率:",recall.mean())
    print("F1_score:",f1_score.mean())
    print("AUC:",auc.mean())



model_name=["lr","tree","svm","forest","Gbdt","Xgbc","gbm"]
for name in model_name:
    model=eval(name)
    print(name)
    muti_score(model)


'''
lr
准确率: 0.7890191148682617
精确率: 0.6542724662896913
召回率: 0.3377975457965613
F1_score: 0.44525012166067884
AUC: 0.7840451024530857
tree
准确率: 0.6962524533638791
精确率: 0.39920670173446693
召回率: 0.4157413593052284
F1_score: 0.40705496051057793
AUC: 0.6029856787858856
svm
准确率: 0.787758390223099
精确率: 0.7351623295760905
召回率: 0.24060335431243626
F1_score: 0.36179547264664874
AUC: 0.7640376541388867
forest
准确率: 0.7921756804332226
精确率: 0.7135700690071172
召回率: 0.2867128441334693
F1_score: 0.40835414886475174
AUC: 0.7752164698827589
Gbdt
准确率: 0.7938590063951863
精确率: 0.6604108594441386
召回率: 0.36633732991104395
F1_score: 0.4708811551285791
AUC: 0.7888240065764295
Xgbc
准确率: 0.7982740847293591
精确率: 0.6829783239831001
召回率: 0.3663162336064133
F1_score: 0.47673826685376613
AUC: 0.7914190511145234
gbm
准确率: 0.79049080811139
精确率: 0.6421783397519263
召回率: 0.3730354066312717
F1_score: 0.47150438344663004
AUC: 0.7776116341798183
'''