scikit-learn機器學習(四)使用決策樹做分類,並畫出決策樹,隨機森林對比

數據來自 UCI 數據集 匹馬印第安人糖尿病數據集

 

載入數據

# -*- coding: utf-8 -*-
import pandas as pd
import matplotlib
matplotlib.rcParams['font.sans-serif']=[u'simHei']
matplotlib.rcParams['axes.unicode_minus']=False
from sklearn.tree import DecisionTreeClassifier
from sklearn.model_selection import train_test_split
from sklearn.metrics import classification_report
from sklearn.pipeline import Pipeline
from sklearn.model_selection import GridSearchCV

from sklearn.datasets import load_breast_cancer

data_set = pd.read_csv('pima-indians-diabetes.csv')
data = data_set.values[:,:]

y = data[:,8]
X = data[:,:8]
X_train,X_test,y_train,y_test = train_test_split(X,y)

 

建立決策樹，網格搜索微調模型

# In[1] 網格搜索微調模型
pipeline = Pipeline([
        ('clf',DecisionTreeClassifier(criterion='entropy'))
        ])
parameters={
        'clf__max_depth':(3,5,10,15,20,25,30,35,40),
        'clf__min_samples_split':(2,3),
        'clf__min_samples_leaf':(1,2,3)
        }
#GridSearchCV 用於系統地遍歷多種參數組合，通過交叉驗證確定最佳效果參數。
grid_search = GridSearchCV(pipeline,parameters,n_jobs=-1,verbose=-1,scoring='f1')
grid_search.fit(X_train,y_train)

# 獲取搜索到的最優參數
best_parameters = grid_search.best_estimator_.get_params()
print("最好的F1值為：",grid_search.best_score_)
print('最好的參數為：')
for param_name in sorted(parameters.keys()):
    print('t%s: %r' % (param_name,best_parameters[param_name]))
    
# In[2] 輸出預測結果並評價
predictions = grid_search.predict(X_test)
print(classification_report(y_test,predictions))

 

最好的F1值為： 0.5573515325670498
最好的參數為：
tclf__max_depth: 5
tclf__min_samples_leaf: 1
tclf__min_samples_split: 2

 

 

 

評價模型

# In[2] 輸出預測結果並評價
predictions = grid_search.predict(X_test)
print(classification_report(y_test,predictions))

 

              precision    recall  f1-score   support

         0.0       0.74      0.89      0.81       124
         1.0       0.67      0.43      0.52        68

 

 

畫出決策樹

# In[3]打印樹
from sklearn import tree  
feature_name=data_set.columns.values.tolist()[:-1]   # 列名稱
DT = tree.DecisionTreeClassifier(criterion='entropy',max_depth=5,min_samples_split=2,min_samples_leaf=5)
DT.fit(X_train,y_train)

'''
# 法一
import pydotplus
from sklearn.externals.six import StringIO
dot_data = StringIO()
tree.export_graphviz(DT,out_file = dot_data,feature_names=feature_name,
                     class_names=["有糖尿病","無病"],filled=True,rounded=True,
                     special_characters=True)
graph = pydotplus.graph_from_dot_data(dot_data.getvalue())
graph.write_pdf("Tree.pdf")
print('Visible tree plot saved as pdf.')
'''

# 法二
import graphviz
#ID3為決策樹分類器fit之后得到的模型，注意這里必須在fit后執行，在predict之后運行會報錯
dot_data = tree.export_graphviz(DT, out_file=None,feature_names=feature_name,class_names=["有糖尿病","無病"]) # doctest: +SKIP
graph = graphviz.Source(dot_data) # doctest: +SKIP
#在同級目錄下生成tree.pdf文件
graph.render("tree2") # doctest: +SKIP

 

 

隨機森林

 

# -*- coding: utf-8 -*-
import pandas as pd
import matplotlib
matplotlib.rcParams['font.sans-serif']=[u'simHei']
matplotlib.rcParams['axes.unicode_minus']=False
from sklearn.tree import DecisionTreeClassifier
from sklearn.model_selection import train_test_split
from sklearn.metrics import classification_report
from sklearn.pipeline import Pipeline
from sklearn.model_selection import GridSearchCV
from sklearn.ensemble import RandomForestClassifier

from sklearn.datasets import load_breast_cancer

data_set = pd.read_csv('pima-indians-diabetes.csv')
data = data_set.values[:,:]

y = data[:,8]
X = data[:,:8]
X_train,X_test,y_train,y_test = train_test_split(X,y)

RF = RandomForestClassifier(n_estimators=10,random_state=11)
RF.fit(X_train,y_train)
predictions = RF.predict(X_test)
print(classification_report(y_test,predictions))

 

              precision    recall  f1-score   support

         0.0       0.82      0.91      0.86       126
         1.0       0.78      0.61      0.68        66

   micro avg       0.81      0.81      0.81       192
   macro avg       0.80      0.76      0.77       192
weighted avg       0.80      0.81      0.80       192