sklearn笔记9 KNN参数cross_val_score调参

完整代码 sklearn代码5 5-KNN参数的筛选

模型包含：
算法
参数

import numpy as np

from sklearn.neighbors import KNeighborsClassifier

from sklearn import datasets

# model_selection：模型选择
# cross_val_score:交叉 validation:验证
# 交叉验证
from sklearn.model_selection import cross_val_score

什么是交叉验证

保留其中一个作为验证数据 其余 作为测试数据，测试完后，取其平均值，是结果更具说服力

凡见到val就可知其中有用于验证的部分

X,y = datasets.load_iris(True)

X.shape

150**0.5    # 样本数量的开平方

# K值选择时从1到13  大概是其开平方的数值，只是用于参考

cross_val_score() 函数的参数选择
Signature: cross_val_score(estimator（算法）, X, y=None, groups=None,
scoring（评分标准）=None, cv=None, n_jobs=1, verbose=0, fit_params=None,
pre_dispatch=‘2*n_jobs’) Docstring: Evaluate a score by
cross-validation

查看评估的标准有哪些

可以到官网上去查看相应的信息
SCikit-learn中文社区

knn = KNeighborsClassifier()
score = cross_val_score(knn,X,y,scoring='accuracy',cv= 10)  #训练了10次
score

knn = KNeighborsClassifier()
score = cross_val_score(knn,X,y,scoring='accuracy',cv= 10)  #训练了10次
score.mean()

errors = []  #误差
for k in range(1,14):
    knn = KNeighborsClassifier(n_neighbors=k)
    
    score = cross_val_score(knn,X,y,scoring='accuracy',cv=6).mean()  # 越接近于1越好
    
#     误差越小 说明K选择越合适
    errors.append(1-score)

import matplotlib.pyplot as plt
%matplotlib inline

plt.plot(np.arange(1,14),errors)

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())

模型如何去筛选最合适参数

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)] =sm
result

取出最大值

np.array(list(result.values())).argmax()

list(result)[20]