如何实现C#中的K均值聚类算法

using MathNet.Numerics.LinearAlgebra;
using MathNet.Numerics.LinearAlgebra.Double;

public class KMeans
{
    private readonly int k; // 聚类数
    private readonly int maxIterations; // 最大迭代次数
    private Matrix<double> data; // 数据
    private Matrix<double> centroids; // 聚类中心

    public KMeans(int k, int maxIterations)
    {
        this.k = k;
        this.maxIterations = maxIterations;
    }

    public void Fit(Matrix<double> data)
    {
        this.data = data;
        Random random = new Random();

        // 随机选择K个数据点作为初始的聚类中心
        centroids = Matrix<double>.Build.Dense(k, data.ColumnCount);
        for (int i = 0; i < k; i++)
        {
            int index = random.Next(data.RowCount);
            centroids.SetRow(i, data.Row(index));
        }

        for (int iteration = 0; iteration < maxIterations; iteration++)
        {
            Matrix<double>[] clusters = new Matrix<double>[k];

            // 初始化聚类
            for (int i = 0; i < k; i++)
            {
                clusters[i] = Matrix<double>.Build.Dense(0, data.ColumnCount);
            }

            // 计算距离并分配数据点到最近的聚类中心
            for (int i = 0; i < data.RowCount; i++)
            {
                Vector<double> point = data.Row(i);
                double minDistance = double.MaxValue;
                int closestCentroid = 0;

                for (int j = 0; j < k; j++)
                {
                    double distance = Distance(point, centroids.Row(j));

                    if (distance < minDistance)
                    {
                        minDistance = distance;
                        closestCentroid = j;
                    }
                }

                clusters[closestCentroid] = clusters[closestCentroid].Stack(point);
            }

            // 更新聚类中心
            for (int i = 0; i < k; i++)
            {
                if (clusters[i].RowCount > 0)
                {
                    centroids.SetRow(i, clusters[i].RowSums().Divide(clusters[i].RowCount));
                }
            }
        }
    }

    private double Distance(Vector<double> a, Vector<double> b)
    {
        return (a.Subtract(b)).Norm(2);
    }
}

public class Program
{
    public static void Main(string[] args)
    {
        Matrix<double> data = Matrix<double>.Build.DenseOfArray(new double[,]
        {
            {1, 2},
            {2, 1},
            {4, 5},
            {5, 4},
            {6, 5},
            {7, 6}
        });

        int k = 2;
        int maxIterations = 100;
        KMeans kMeans = new KMeans(k, maxIterations);
        kMeans.Fit(data);

        // 输出聚类结果
        Console.WriteLine("聚类中心：");
        Console.WriteLine(kMeans.Centroids);
    }
}