交叉驗證

導包

import numpy as npfrom sklearn.neighbors import KNeighborsClassifierfrom sklearn import datasets#model_selection ：模型選擇
# cross_val_score: 交叉 ，validation：驗證（測試）
#交叉驗證
from sklearn.model_selection import cross_val_score

讀取datasets中鳶尾花(yuan1wei3hua)數據

X,y= datasets.load_iris(True)
X.shape

(150, 4)

一般情況不會超過數據的開方數

#參考
150**0.5
#K 選擇 1~13

12.24744871391589

knn = KNeighborsClassifier()score = cross_val_score(knn,X,y,scoring='balanced_accuracy',cv=11)
score.mean()

0.968181818181818

應用cross_val_score篩選 n_neighbors k值

errors =[]
for k in range(1,14):knn = KNeighborsClassifier(n_neighbors=k)score = cross_val_score(knn,X,y, scoring='accuracy',cv = 6).mean()#誤差越小 說明K選擇越合適 越好errors.append(1-score)import matplotlib.pyplot as plt
%matplotlib inline#k = 11時 誤差最小 說明最合適的k值
plt.plot(np.arange(1,14),errors)

[<matplotlib.lines.Line2D at 0x17ece9ff0b8>]

應用cross_val_score篩選 weights

weights =['uniform','distance']for w  in weights:knn  = KNeighborsClassifier(n_neighbors = 11,weights= w)print(w,cross_val_score(knn,X,y, scoring='accuracy',cv = 6).mean())

uniform 0.98070987654321
distance 0.9799382716049383

模型如何調參的，參數調節

result = {}
for k in range(1,14):for w  in weights:knn = KNeighborsClassifier(n_neighbors=k,weights=w)sm = cross_val_score(knn,X,y,scoring='accuracy',cv=6).mean()result[w+str(k)] =sma =result.values()
list(a)np.array(list(a)).argmax()

20

list(result)[20]

‘uniform11’