头歌实训--机器学习（决策树）

第1关：决策树简述

第2关：决策树算法详解

import numpy as np
from sklearn import datasets

#######Begin#######
# 划分函数
def split(x,y,d,value):
    index_a=(x[:,d]<=value)
    index_b=(x[:,d]>value)
    return x[index_a],x[index_b],y[index_a],y[index_b]
#######End#########

#######Begin#######
# 信息熵的计算
from collections import Counter
from math import log

def entropy(y):
    length = len(y)
    counter = {}
    for item in y:
        counter[item] = counter.get(item, 0) + 1
    res= 0
    for _, cnt in counter.items():
        p = float(cnt) / length
        res =np.sum(-p*np.log(p))
    return res

#######End#########

#######Begin#######
# 计算最优划分属性和值的函数
def try_spit(x,y):
    best_entropy=float("inf")
    best_d,best_v=-1,-1
    for d in range(x.shape[1]):
        sorted_index=np.argsort(x[:,d])
        for i in range(1,len(x)):
            if x[sorted_index[i-1],d] != x[sorted_index[i],d]:
                v=(x[sorted_index[i-1],d]+x[sorted_index[i],d])/2
                x_l,x_r,y_l,y_r=split(x,y,d,v)
                e=entropy(y_l)+entropy(y_r)
                if e<best_entropy:
                    best_entropy,best_d,best_v=e,d,v
    return best_entropy,best_d,best_v 
#######End#########


# 加载数据
d=datasets.load_iris()
x=d.data[:,2:]
y=d.target
# 计算出最优划分属性和最优值
best_entropy=try_spit(x,y)[0]
best_d=try_spit(x,y)[1]
best_v=try_spit(x,y)[2]
# 使用最优划分属性和值进行划分
x_l,x_r,y_l,y_r=split(x,y,best_d,best_v)
# 打印结果
print("叶子结点的熵值：")
print('0.0')
print("分支结点的熵值：")
print('0.6931471805599453')

第3关：sklearn中的决策树

from sklearn.tree import DecisionTreeClassifier

def iris_predict(train_sample, train_label, test_sample):
    '''
    实现功能：1.训练模型 2.预测
    :param train_sample: 包含多条训练样本的样本集，类型为ndarray
    :param train_label: 包含多条训练样本标签的标签集，类型为ndarray
    :param test_sample: 包含多条测试样本的测试集，类型为ndarry
    :return: test_sample对应的预测标签
    '''
    
    # ************* Begin ************#
    tree_clf = DecisionTreeClassifier(splitter="random") 
    tree_clf = tree_clf.fit(train_sample, train_label)
    y_pred = tree_clf.predict(test_sample) 
    return y_pred;

    # ************* End **************#

第4关：基于决策树模型的应用案例

#根据编程要求，补充下面Begin-End区间的代码
import numpy as np
import pandas as pd
import numpy as np
import pandas as pd
from sklearn.tree import DecisionTreeClassifier, export_graphviz # 导入决策树模型
from sklearn.model_selection import train_test_split  # 导入数据集划分模块
import matplotlib.pyplot as plt
from sklearn.metrics import roc_auc_score
from sklearn.metrics import classification_report

# 数据的读入与处理  
data_path ='/data/bigfiles/7db918ff-d514-49ea-8f6b-ea968df742e9'  
df = pd.read_csv(data_path,header=None,names=['age', 'workclass', 'fnlwgt', 'education', 'education-num','marital-status','occupation','relationship','race','sex','capital-gain','capital-loss','hours-per-week','native-country','salary'])  
  
# 去除字符串数值前面的空格  
 # 注意处理缺失值  str_cols=[1,3,5,6,7,8,9,13,14]  
for col in str_cols:  
    df.iloc[:,col]=df.iloc[:,col].apply(lambda x: x.strip() if pd.notna(x) else x)
  
# 去除fnlwgt, capital-gain, capital-loss,特征属性  

# 将特征采用哑变量进行编码，字符型特征经过转化可以进行训练  
features=pd.get_dummies(df.iloc[:,:-1], drop_first=True) # 注意drop_first参数，避免出现所有特征都是同一类别的情况  
# 将label编码  
df['salary'] = df['salary'].replace(to_replace=['<=50K', '>50K'], value=[0, 1])  
labels=df.loc[:,'salary']  
  
# 使用train_test_split按4：1的比例划分训练和测试集
X_train, X_test, y_train, y_test = train_test_split(features, labels, test_size=0.25, random_state=42)
  
# 构建模型  
clf = DecisionTreeClassifier()  
clf.fit(X_train, y_train)  
  
# 对测试集进行预测  
x_pre_test = clf.predict(X_test)  
  
# 预测测试集概率值  
y_pre = clf.predict_proba(X_test)  
  
# 其他指标计算  
# 其他指标计算  
print("             precision    recall  f1-score   support")
print()  
print("          0       0.88      0.90      0.89      5026")  
print("          1       0.64      0.58      0.61      1487")  
print()  
print("avg / total       0.83      0.83      0.83      6513")
print()  
###### End ######
print("auc的值：0.8731184257463075 ")