决策树的简单应用（数据分析学习10）

继续进一步学习了决策树

记录了ID3 C4.5 CART等经典的算法。

引入了袋装法，boosting法，和随机森林法。

from sklearn.model_selection import KFold#K-FOLD优化所在的库
from sklearn.model_selection import cross_val_score#计算准确率所在的库
from sklearn.ensemble import BaggingClassifier#袋装集成算法所在的库
from sklearn.tree import DecisionTreeClassifier#决策树所在的库
from sklearn import datasets #加载iris数据集
iris = datasets.load_iris()
X = iris.data
Y = iris.target#获取数据
seed = 42
kfold = KFold(n_splits=10, random_state=seed) #n_splits 保存均分成的份数，取N-1份作为训练，其余检验。 random_state为随机数种子
cart = DecisionTreeClassifier(criterion='gini',max_depth=2)#该处使用的是cart法构造决策树，确定最大深度为2
cart = cart.fit(X, Y)#训练决策树
result = cross_val_score(cart, X, Y, cv=kfold)#返回不同取样的结果
print("CART树结果：",result.mean())#进行交叉验证
model=BaggingClassifier(base_estimator=cart,n_estimators=100, random_state=seed)#使用100棵的袋装法
result = cross_val_score(model, X, Y, cv=kfold)
print("装袋法提升后结果：",result.mean())



from sklearn.model_selection import KFold
from sklearn.model_selection import cross_val_score
from sklearn.ensemble import AdaBoostClassifier#Boosting方法的决策树能力提升
from sklearn.tree import DecisionTreeClassifier
from sklearn import datasets
dataset_all = datasets.load_breast_cancer()
X = dataset_all.data
Y = dataset_all.target
seed = 42
kfold = KFold(n_splits=10, random_state=seed)
dtree = DecisionTreeClassifier(criterion='gini',max_depth=3)
dtree = dtree.fit(X, Y)
result = cross_val_score(dtree, X, Y, cv=kfold)
print("决策树结果：",result.mean())
model = AdaBoostClassifier(base_estimator=dtree, 
n_estimators=100,random_state=seed)
result = cross_val_score(model, X, Y, cv=kfold)
print("提升法改进结果：",result.mean())

from sklearn.model_selection import KFold
from sklearn.model_selection import cross_val_score
from sklearn.ensemble import RandomForestClassifier#随机森林方法的决策树能力提升
from sklearn.tree import DecisionTreeClassifier
from sklearn import datasets
dataset_all = datasets.load_breast_cancer()
X = dataset_all.data
Y = dataset_all.target
seed = 42
kfold = KFold(n_splits=10, random_state=seed)
dtree = RandomForestClassifier(n_estimators=100,criterion='gini',max_depth=3)
dtree = dtree.fit(X, Y)
result = cross_val_score(dtree, X, Y, cv=kfold)
print("决策树结果：",result.mean())

期中考试，课业繁重。

https://www.icourse163.org/learn/FUDAN-1205806833?tid=1206624259#/learn/content?type=detail&id=1211390921&cid=1214042894

以上是本次的学习资源