from sklearn.tree import DecisionTreeClassifier
前篇使用了
K-近鄰演算法(K Nearest Neighbor ,簡稱 KNN)
本篇要改用決策樹分類
處理的資料: https://pse.is/3ty6rk
部分資料:
import pandas as pd
from sklearn.neighbors import KNeighborsClassifier
from sklearn.model_selection import train_test_split
folder = “C:\Python\P107\doc”
fname = “student.csv”
import os
fpath = os.path.join(folder,fname)
# fpath = folder + “\\” + fname #同義
df = pd.read_csv(fpath,header=None,skiprows=[0])
df.to_excel(os.path.join(folder,”knsDF.xlsx”))
#df.index.size = 40 #df.columns.size = 5
X = df.drop([4],axis=1).values
y = df[4].values
Xtrain, Xtest, ytrain, ytest =\
train_test_split(X,y,test_size=0.25,
random_state= 42 ,shuffle =True)
XtestDF = pd.DataFrame(Xtest)
XtestDF.to_excel(os.path.join(folder,”knsXtest.xlsx”))
print(“Xtrain.shape:”,Xtrain.shape) #(30, 3)
print(“ytrain.shape:”,ytrain.shape) #(30,)
print(“Xtest.shape:”,Xtest.shape) #(10, 3)
print(“ytest.shape:”,ytest.shape) #(10,)
from sklearn.tree import DecisionTreeClassifier
for cri in [“gini”,”entropy”]:
tree = DecisionTreeClassifier(criterion = cri)
tree.fit(Xtrain,ytrain)
pred = tree.predict(Xtest)
pred2 = tree.predict_proba(Xtest)
print(“prediction:”,pred,”in case criterion”,cri )
print(“prediction proability:”,pred2,
“\nin case criterion”,cri )
howgood = tree.score(Xtest,ytest)
print(“Goodness:”,howgood,”in case ctrterion”,cri)
#前半部同KNN
後半部類似的語法,
主要改用 DecisionTreeClassifier
輸出結果:
就算有第0欄(index,非資料)干擾
決策樹的score仍比
匯出決策樹:
改寫後半段程式碼:
from sklearn.tree import DecisionTreeClassifier
from sklearn.tree import export_text
for cri in [“gini”,”entropy”]:
tree = DecisionTreeClassifier(criterion = cri)
tree.fit(Xtrain,ytrain)
r = export_text(tree,feature_names =
[“item”,”English”,”Math”,”Chinese”],
show_weights=True)
print(r)
輸出結果:
criterion = “gini”
criterion = “entropy”
推薦hahow線上學習python: https://igrape.net/30afN