手动实现k-means聚类使用noise_moon数据集

以下是手动实现k-means聚类使用noise_moon数据集的Python代码：

import numpy as np
import matplotlib.pyplot as plt
from sklearn.datasets import make_moons

# 生成noise_moon数据集
X, y = make_moons(n_samples=200, noise=0.05, random_state=0)

# 定义k-means聚类函数
def k_means(X, k, max_iter=1000):
    n_samples, n_features = X.shape

    # 随机初始化k个聚类中心
    centroids = X[np.random.choice(n_samples, k, replace=False)]

    # 迭代更新聚类中心和聚类标签
    for i in range(max_iter):
        distances = np.zeros((n_samples, k))
        for j in range(k):
            distances[:, j] = np.linalg.norm(X - centroids[j], axis=1)
        labels = np.argmin(distances, axis=1)
        new_centroids = np.zeros((k, n_features))
        for j in range(k):
            new_centroids[j] = np.mean(X[labels == j], axis=0)
        if np.all(centroids == new_centroids):
            break
        centroids = new_centroids

    return centroids, labels

# 对数据集进行聚类
k = 2
centroids, labels = k_means(X, k)

# 绘制聚类结果
colors = ['blue', 'red']
for i in range(k):
    plt.scatter(X[labels == i, 0], X[labels == i, 1], c=colors[i])
plt.scatter(centroids[:, 0], centroids[:, 1], marker='*', s=200, c='black')
plt.title('k-means Clustering with k=%d' % k)
plt.show()

上述代码生成了一个`noise_moon`数据集，并定义了一个`k-means`聚类函数。通过调用`k_means`函数对数据集进行聚类，并绘制聚类结果。结果如下图所示：