kmeans算法

基本名词解释

算法根据有没有标签分为 监督学习算法和无监督学习的算法
kmeans ：没有标签 数据无监督学习的聚类算法
聚类算法 ：就是把相似的东西分到一组
kmeans 中k 表示分组的个数 （族的个数）
质心： 每个族中的数据各个向量的平均值
距离的度量 ：欧几里德距离 和余弦相似度（数据需要归一化处理）
优化目标 $$\sum _{i=1}^k\sum _{xϵci}dist(ci,x{)}^2$$

训练过程
1根据k值随机设置质点的坐标
2遍历所有样本到质点的距离 分组
3重新根据每组数据的坐标更换质点的坐标
重复 步骤2 3

完整代码

import numpy as np

class KMeans:
    def __init__(self,data,num_clustres):
        self.data = data
        self.num_clustres = num_clustres
        
    def train(self,max_iterations):
        #1.先随机选择K个中心点
        centroids = KMeans.centroids_init(self.data,self.num_clustres)
        #2.开始训练
        num_examples = self.data.shape[0]
        closest_centroids_ids = np.empty((num_examples,1))
        for _ in range(max_iterations):
            #3得到当前每一个样本点到K个中心点的距离，找到最近的
            closest_centroids_ids = KMeans.centroids_find_closest(self.data,centroids)
            #4.进行中心点位置更新
            centroids = KMeans.centroids_compute(self.data,closest_centroids_ids,self.num_clustres)
        return centroids,closest_centroids_ids
                
    @staticmethod    
    def centroids_init(data,num_clustres):
        num_examples = data.shape[0]
        random_ids = np.random.permutation(num_examples)
        centroids = data[random_ids[:num_clustres],:]
        return centroids
    @staticmethod 
    def centroids_find_closest(data,centroids):
        num_examples = data.shape[0]
        num_centroids = centroids.shape[0]
        closest_centroids_ids = np.zeros((num_examples,1))
        for example_index in range(num_examples):
            distance = np.zeros((num_centroids,1))
            for centroid_index in range(num_centroids):
                distance_diff = data[example_index,:] - centroids[centroid_index,:]
                distance[centroid_index] = np.sum(distance_diff**2)
            closest_centroids_ids[example_index] = np.argmin(distance)
        return closest_centroids_ids
    @staticmethod
    def centroids_compute(data,closest_centroids_ids,num_clustres):
        num_features = data.shape[1]
        centroids = np.zeros((num_clustres,num_features))
        for centroid_id in range(num_clustres):
            closest_ids = closest_centroids_ids == centroid_id
            centroids[centroid_id] = np.mean(data[closest_ids.flatten(),:],axis=0)
        return centroids