K近邻

本案例使数据：官方数据

语言：python

工具：Jupyter

一、导包

# 导包
from sklearn.neighbors import KNeighborsClassifier
from sklearn.datasets import load_wine
from sklearn.model_selection import train_test_split
import sklearn.metrics as sm

二、导入数据

# 导入数据
data = load_wine()

三、查看数据

# 查看数据，赋值
X, y = data.data, data.target

四、划分数据集

# 划分数据集
train_x, test_x, train_y, test_y = train_test_split(X, y, train_size=0.8, random_state=42)

五、创建模型

# 创建模型
model = KNeighborsClassifier(n_neighbors=3)

六、训练数据

# 训练数据
model.fit(train_x, train_y)KNeighborsClassifier(n_neighbors=3)

七、预测

# 预测
pred_test_y = model.predict(test_x)pred_test_yarray([2, 0, 2, 0, 1, 0, 1, 2, 0, 0, 2, 2, 0, 1, 0, 1, 1, 1, 0, 1, 0, 1,2, 2, 1, 2, 1, 2, 1, 0, 0, 1, 2, 0, 0, 0])

八、模型评估

1.准确率

# 1.准确率
accuracy = sm.accuracy_score(test_y, pred_test_y)accuracy
0.8055555555555556

2、精确率

# 2.精确率
precision = sm.precision_score(test_y,pred_test_y, average='micro')precision
0.8055555555555556

3.召回率

# 3.召回率
recall = sm.recall_score(test_y,pred_test_y, average='macro')recall
0.7976190476190476

4.f1_score

# 4. f1_score
f1 = sm.f1_score(test_y, pred_test_y,average='macro')f1
0.7899877899877898

5.混淆矩阵

# 5. 混淆矩阵
sm.confusion_matrix(test_y,pred_test_y)
array([[12,  0,  2],[ 1, 11,  2],[ 1,  1,  6]], dtype=int64)

6.分类报告

# 6.分类报告
print(sm.classification_report(test_y,pred_test_y))precision    recall  f1-score   support0       0.86      0.86      0.86        141       0.92      0.79      0.85        142       0.60      0.75      0.67         8accuracy                           0.81        36macro avg       0.79      0.80      0.79        36
weighted avg       0.82      0.81      0.81        36