Python_sklearn Machine Learning Library Learning notes (iv) Decision_tree (decision Tree)

# Decision Tree

Import pandas as Pdfrom sklearn.tree import decisiontreeclassifierfrom sklearn.cross_validation import Train_test_ Splitfrom sklearn.metrics Import classification_reportfrom sklearn.pipeline import Pipelinefrom Sklearn.grid_search Import GRIDSEARCHCV

Import zipfile# compression saves space z=zipfile. ZipFile (' Ad-dataset.zip ') # df=pd.read_csv (Z.open (z.namelist () [0]), header=none,low_memory=false) # df = Pd.read_csv ( Z.open (z.namelist () [0]), Header=none, Low_memory=false)

Df=pd.read_csv ('. \\tree_data\\ad.data ', Header=none) Explanatory_variable_columns=set (df.columns.values) response _variable_column=df[len (Df.columns.values)-1] #最后一列是代表的标签类型explanatory_variable_columns. Remove (len (df.columns)- 1)

y=[1 if E = = ' ad. ' Else 0 for E in Response_variable_column]x=df.loc[:,list (Explanatory_variable_columns)]

#匹配? characters, and convert the value to -1x.replace (to_replace= ' *\? ', Value=-1, Regex=true, Inplace=true)

X_train,x_test,y_train,y_test=train_test_split (x, y) #用信息增益启发式算法建立决策树pipeline =pipeline ([' CLF ', Decisiontreeclassifier (criterion= ' entropy ')]) parameters = {' clf__max_depth ': [155], Clf__min_samples_ Split ': (1, 2, 3), ' Clf__min_samples_leaf ': (1, 2, 3)} #f1查全率和查准率的调和平均grid_search =GRIDSEARCHCV (pipeline,parameters,n_ Jobs=-1,                         verbose=1,scoring= ' F1 ') grid_search.fit (x_train,y_train) print ' best results:%0.3f '%grid_search.best_score_ print ' Optimal parameters ' Best_parameters=grid_search.best_estimator_.get_params () best_parameters

Output Result:

Fitting 3 folds for each of the candidates, totalling Bayi fits

[Parallel (N_jobs=-1)]: Done to  tasks      | elapsed:   21.0s[parallel (n_jobs=-1)]: Done  Bayi out of  81 | Elapsed:   34.7s finished

Best results: 0.888 optimal parameters

OUT[123]:

{' CLF ': Decisiontreeclassifier (Class_weight=none, criterion= ' entropy ', max_depth=160,             Max_features=none, Max_ Leaf_nodes=none, Min_samples_leaf=1,             min_samples_split=3, min_weight_fraction_leaf=0.0,             Presort=False, Random_state=none, splitter= ' best '), ' clf__class_weight ': None, ' clf__criterion ': ' Entropy ', ' clf__max_depth ': 160, ' Clf__max_features ': None, ' clf__max_leaf_nodes ': None, ' clf__min_samples_leaf ': 1, ' Clf__min_samples_split ': 3, ' clf__ Min_weight_fraction_leaf ': 0.0, ' clf__presort ': False, ' clf__random_state ': None, ' clf__splitter ': ' Best ', ' Steps ': [(' CLF ',   decisiontreeclassifier (class_weight=none, criterion= ' entropy ', max_depth=160,               Max_features=none, Max_leaf_nodes=none, Min_samples_leaf=1,               min_samples_split=3, min_weight_fraction_leaf=0.0,               presort= False, Random_state=none, splitter= ' best ')]}

For Param_name in Sorted (Parameters.keys ()):    print (' \t%s:%r '% (Param_name,best_parameters[param_name])) Predictions=grid_search.predict (x_test) print Classification_report (y_test,predictions)

Output Result:

clf__max_depth:150
Clf__min_samples_leaf:1
Clf__min_samples_split:1
Precision Recall F1-score Support

0 0.97 0.99) 0.98 703
1 0.91 0.84) 0.87 117

Avg/total 0.96 0.96 0.96 820

Df.head ()

Output results;

9

	0	1	2	3	4	5	6	7	8		...	1549	1550	1551	1552	1553	1554	1555	1556	1557	1558
0	125	125	1.0	1	0	0	0	0	0	0	...	0	0	0	0	0	0	0	0	0	Ad.
1	57	468	8.2105	1	0	0	0	0	0	0	...	0	0	0	0	0	0	0	0	0	Ad.
2	33	230	6.9696	1	0	0	0	0	0	0	...	0	0	0	0	0	0	0	0	0	Ad.
3	60	468	7.8	1	0	0	0	0	0	0	...	0	0	0	0	0	0	0	0	0	Ad.
4	60	468	7.8	1	0	0	0	0	0	0	...	0	0	0	0	0	0	0	0	0	Ad.

# Decision Tree Integration

#coding: Utf-8import Pandas as Pdfrom sklearn.ensemble import Randomforestclassifierfrom sklearn.cross_validation Import train_test_splitfrom sklearn.metrics import classification_reportfrom sklearn.pipeline import Pipelinefrom Sklearn.grid_search Import Gridsearchcvdf=pd.read_csv ('. \\tree_data\\ad.data ', Header=none,low_memory=false) Explanatory_variable_columns=set (df.columns.values) Response_variable_column=df[len (df.columns.values)-1]

Df.head ()

9

	0	1	2	3	4	5	6	7	8		...	1549	1550	1551	1552	1553	1554	1555	1556	1557	1558
0	125	125	1.0	1	0	0	0	0	0	0	...	0	0	0	0	0	0	0	0	0	Ad.
1	57	468	8.2105	1	0	0	0	0	0	0	...	0	0	0	0	0	0	0	0	0	Ad.
2	33	230	6.9696	1	0	0	0	0	0	0	...	0	0	0	0	0	0	0	0	0	Ad.
3	60	468	7.8	1	0	0	0	0	0	0	...	0	0	0	0	0	0	0	0	0	Ad.
4	60	468	7.8	1	0	0	0	0	0	0	...	0	0	0	0	0	0	0	0	0	Ad.

#The Last column describes the targets (remove the final row) Explanatory_variable_columns.remove (Len (df.columns.values)-1) y=[1 if e= = ' ad. ' Else 0 for E in Response_variable_column]x=df.loc[:,list (explanatory_variable_columns)] #置换有? For -1x.replace (to_replace= ' *\? ', Value=-1, Regex=true, inplace=true) x_train,x_test,y_train,y_test=train_test_split (x, Y) pipeline=pipeline ([' CLF ', randomforestclassifier (criterion= ' entropy ')]) parameters = {' Clf__n_estimators ': ( 5, "Clf__max_depth": (1, 2, 3), ' Clf__min_samples_leaf ': (+ 1, 2, 3), ' Clf__min_samples_split ': D_search = GRIDSEARCHCV (pipeline,parameters,n_jobs=-1,verbose=1,scoring= ' F1 ') Grid_search.fit (X_train,y_train)

Print (U ' best effect:%0.3f '%grid_search.best_score_) print U ' optimal parameters: ' best_parameters=grid_search.best_estimator_.get_ params () for param_name in sorted (Parameters.keys ()):    print (' \t%s:%r '% (Param_name,best_parameters[param_name]) )

Output Result:

Best results: 0.929 optimal parameters: clf__max_depth:250 clf__min_samples_leaf:1 clf__min_samples_split:3 clf__n_estimators:50

Predictions=grid_search.predict (x_test) print Classification_report (y_test,predictions)

Output Result:

Precision Recall F1-score Support

0 0.98 1.00) 0.99 705
1 0.97 0.90) 0.93 115

Avg/total 0.98 0.98 0.98 820

Python_sklearn Machine Learning Library Learning notes (iv) Decision_tree (decision tree)