23、深度学习之决策树学习

具体参考决策树基本原理和讲解:

信息熵的学习参考:链接:https://pan.baidu.com/s/1_AF7xoUhm3XgcYGfGrk5ng 
提取码:kztu

决策树的学习参考:https://sklearn.apachecn.org/docs/master/11.html

代码使用的csv参考:链接:https://pan.baidu.com/s/1IxbFuwET7qMbJyEu76WROg 
提取码:g7f7

决策树代码

from sklearn import tree
from sklearn.feature_extraction import DictVectorizer
from sklearn import preprocessing
import pandas as pb
import glob
import numpy as np
file_csv=glob.glob(r"F:\DTree\*.csv")
total_decison=[]
list_decision=[]
label_decision=[]
for csv in file_csv:
    df=pb.read_csv(csv,index_col=False)
    for indexs in df.index:
        if indexs==0:
            headInfo = list(df.head(indexs))[1:-1]
            continue
        label_decision.append(df.iloc[indexs,-1])
        list_decision.append(list(df.iloc[indexs,1:-1]))

    for item in list_decision:
        dict_dec = {}
        for index,it in enumerate(item):
           dict_dec[headInfo[index]]=it
        total_decison.append(dict_dec)
print(total_decison)
print(label_decision)

vec=DictVectorizer()

dumpx=vec.fit_transform(total_decison).toarray()
print("dumpx",str(dumpx))

lb=preprocessing.LabelBinarizer()
dumpy=lb.fit_transform(label_decision)
print("dumpy",str(dumpy))

clf=tree.DecisionTreeClassifier(criterion='entropy')
clf=clf.fit(dumpx,dumpy)
print(str(clf))

with open("F:/a.dot","w") as f:
    f=tree.export_graphviz(clf,feature_names=vec.get_feature_names(),out_file=f)


newRowX = dumpx[0, :]
print("oneRowX: " + str(newRowX))


newRowX[0] = 1
newRowX[2] = 0
print("newRowX: " + str(newRowX))

predictedY = clf.predict(np.array(newRowX).reshape(1, -1))
print("predictedY: " + str(predictedY))

 

上一篇:常用的异常检测代码


下一篇:【笔记】KNN之网格搜索与k近邻算法中更多超参数