利用决策树进行企鹅分类

数据集：

数据集获取，百度网盘：

复制这段内容后打开百度网盘手机App，操作更方便哦 链接:

网盘地址

https://pan.baidu.com/s/1i9PAcOQ9g15qtMyiCvl3lQ

提取码:6z30

上代码：

import matplotlib.pyplot as plt
import pandas as pd
from sklearn.tree import DecisionTreeClassifier
import seaborn as sns
import numpy as np
# 前三个函数是为了把.xlsx文件转化为.csv
def get_excel_sheet_list(filename):
    df = pd.read_excel(filename, sheet_name=None,engine='openpyxl')
    return df.keys()

def get_excel_sheet_data(filename, sheetname=None):
    df = pd.read_excel(filename, sheet_name=sheetname,engine='openpyxl')
    dataList = df.to_dict(orient='records')  # 转换为列表
    return dataList

def file_trans(Read_filepath):
    sheet_list = get_excel_sheet_list(Read_filepath)
    file_list=[]
    for file in sheet_list:
        Save_filepath = file + '.csv'
        file_list.append(Save_filepath)
        sheetData = pd.DataFrame(get_excel_sheet_data(Read_filepath, file))
        if sheetData.empty:
            continue
        sheetData.to_csv(Save_filepath, index=False, mode='w', encoding='utf-8')
    return file_list

# 补足缺省值和返回训练数据，切片
def getDataSet():
    Read_filepath = "../xiangmu/pythonProject/数据集_25.xlsx"
    list_file = file_trans(Read_filepath)
    df = pd.read_csv(list_file[0])
    # 所用的数据
    df_usage=df.iloc[:,:]
    df_usage.columns=['年龄','阶段','地区','文化程度','家庭教养','孩子性格','家庭情感','孩子应对','自我效能','交流情况','上网目的','创伤成长']
    df_index = df_usage.columns.values
    # df_usage = df_usage.reindex(np.random.permutation(df_usage.index))
    # print(df_index)
    # 数据转化为列表
    df_list=df_usage.values.tolist()
    return df_list,df_index

def splitData(aaaaaaaaaa,bbbbbbbbbb):
    nrow = len(aaaaaaaaaa)
    ncol = len(bbbbbbbbbb)
    nxval=5
    test_data=[]
    train_data=[]
    for ixval in range(nxval):
        # 每一次循环将count置为零
        test_index = [a for a in range(nrow) if a % nxval == ixval]
        train_index = [a for a in range(nrow) if a % nxval != ixval]
        test_data = [aaaaaaaaaa[r] for r in test_index]
        train_data = [aaaaaaaaaa[r] for r in train_index]
    x_train=[]
    y_train=[]
    x_test=[]
    y_test=[]
    for i in test_data:
        x_test.append(i[:-1])
        y_test.append(i[-1])
    for j in train_data:
        x_train.append(j[:-1])
        y_train.append(j[-1])
    return x_train,y_train,x_test,y_test

aa,bb=getDataSet()
x_train,y_train,x_test,y_test=splitData(aa,bb)
## 定义 决策树模型
clf = DecisionTreeClassifier(criterion='entropy',max_depth=10,min_samples_leaf=2,min_samples_split=2)
# 在训练集上训练决策树模型
clf.fit(x_train, y_train)

## 在训练集和测试集上分布利用训练好的模型进行预测
train_predict = clf.predict(x_train)
test_predict = clf.predict(x_test)
from sklearn import metrics

## 利用accuracy（准确度）【预测正确的样本数目占总预测样本数目的比例】评估模型效果
print('训练准确率:',metrics.accuracy_score(y_train,train_predict))
print('测试准确率:',metrics.accuracy_score(y_test,test_predict))

## 查看混淆矩阵 (预测值和真实值的各类情况统计矩阵)
confusion_matrix_result = metrics.confusion_matrix(test_predict,y_test)
print('混淆矩阵:\n',confusion_matrix_result)

# 利用热力图对于结果进行可视化
plt.figure(figsize=(8, 6))
sns.heatmap(confusion_matrix_result, annot=True, cmap='BrBG_r')
plt.xlabel('true')
plt.ylabel('predict')
plt.show()

结果：

 

 

准确率达到0.97，已经是很高的数值了