决策树

import pandas as pd
import os
from sklearn.tree import DecisionTreeClassifier as DTC
from sklearn.tree import export_graphviz
BASE_DIR = os.path.dirname(os.path.abspath(__file__))
fname = os.path.join(BASE_DIR, 'data', 'lesson.xlsx')
df = pd.read_excel(fname, encoding='utf-8')

x = df[['实战', '课时数', '是否促销', '是否提供配套资料']].values
y = df[['销量']].values

for i in range(0, len(x)):
    for j in range(0, len(x[i])):
        cur = x[i][j]
        if cur == '是' or cur == '高' or cur == '多':
            x[i][j] = 1
        else:
            x[i][j] = -1

for i in range(0, len(y)):
    cur = y[i]
    if cur == '高':
        y[i] = 1
    else:
        y[i] = -1

# 建立决策树
dtc=DTC(criterion='entropy')
dtc.fit(x, y.astype('int'))

# 可视化决策树
with open(os.path.join(BASE_DIR, 'data', 'dtc.dot'), 'w') as f:
  export_graphviz(dtc, feature_names=['shizhan', 'keshi', 'cuxiao', 'ziliao'], out_file=f)

# 下载graphviz软件
# 添加graphviz环境变量
# 进入到dot格式文件目录cmd执行：dot -Tpng dtc.dot -o lesson.png
# 决策树往左看负能量
# entropy信息熵，信息熵越大越可靠，越小越不可靠
# sample总的信息条数
# values消息高低的统计情况