决策树--简单学习

从sklearn中导入决策树分类器。

import pandas as pd 
import numpy as np
from sklearn.model_selection import train_test_split ,cross_val_score
from sklearn.tree import DecisionTreeClassifier 
from sklearn.datasets import load_wine
#下载数据
wine = load_wine()
#转为Dataframe

data_wine = pd.DataFrame(wine.data,columns=wine.feature_names)
data_wine["y"] = wine.target

#划分训练集与测试集
xtrain,xtest,ytrain,ytest = train_test_split(data_wine.iloc[:,:-1],data_wine.y,test_size=0.3,random_state=6)
#拟合训练集
dtc = DecisionTreeClassifier(max_depth=4,
                             random_state=2)
dtc_wine = dtc.fit(xtrain,ytrain) 
dtc.score(xtest,ytest)                           

score为0.9444444444444444。

绘制决策树:

import graphviz
from sklearn import tree

wine_data2 = tree.export_graphviz(dtc_wine,
                                 out_file= None,
                                 feature_names=feature_name,
                                 class_names=["朗姆","琴酒","伏特加"],
                                 filled= True,
                                 rounded=True)
graph = graphviz.Source(wine_data2)
graph

绘图如下:

图形可视化graphviz，需要下载，下载地址：
https://www.graphviz.org/download/