再测试Wine数据集

为了测试Wine数据集，我们可以再次使用K-means聚类算法。Wine数据集包含三种葡萄酒的数据，每种都有13个特征。让我们先加载数据，然后按照类似步骤进行操作：

1. 导入必要的库和数据集

import pandas as pd
from sklearn.datasets import load_wine
from sklearn.model_selection import train_test_split
from sklearn.preprocessing import StandardScaler
from sklearn.cluster import KMeans
import seaborn as sns
import matplotlib.pyplot as plt

wine_data = load_wine()
df_wine = pd.DataFrame(data=wine_data.data, columns=wine_data.feature_names)
target = wine_data.target

2. 划分训练集和测试集

X_train, X_test, y_train, y_test = train_test_split(df_wine, target, test_size=0.2, random_state=42)

3. 数据预处理

scaler = StandardScaler()
X_train_scaled = scaler.fit_transform(X_train)
X_test_scaled = scaler.transform(X_test)

4. 选择K值（这里同样假设3个类别），创建并训练KMeans模型

n_clusters = 3
kmeans = KMeans(n_clusters=n_clusters, random_state=42)
kmeans.fit(X_train_scaled)

5. 预测和评估

labels_train = kmeans.predict(X_train_scaled)
labels_test = kmeans.predict(X_test_scaled)

# 可视化聚类结果
sns.pairplot(df_wine, hue=labels_train, diag_kind="kde")
plt.show()

# 计算训练集和测试集的轮廓系数
train_score = silhouette_score(X_train_scaled, labels_train)
test_score = silhouette_score(X_test_scaled, labels_test)
print(f"训练集轮廓系数: {train_score}")
print(f"测试集轮廓系数: {test_score}")

通过观察散点图和轮廓系数，你可以评估K-means聚类效果是否良好，以及是否存在过拟合或欠拟合的问题。