聚类实验

运行一个聚类算法； 要求： 1) 数据自选，例如：iris 2) 修改给出的代码，运行得到结果，例如修改类别、修改算法参数、更换聚类算法等；3）打印出评价指标；

K-means实验代码

import matplotlib.pyplot as plt
import numpy as np
from sklearn.cluster 
import KMeansfrom sklearn 
import datasets
from sklearn.metrics import fowlkes_mallows_scoreimport sklearn

iris = datasets.load_iris()
X = iris.data[:, :4]  # #表示我们取特征空间中的4个维度

estimator = KMeans(n_clusters=3)  # 构造聚类器estimator.fit(X)  # 聚类
label_pred = estimator.labels_  # 获取聚类标签 
 # DBI的值最小是0，值越小，代表聚类效果越好
FMI =fowlkes_mallows_score(iris.data[:,-1], label_pred)
print("FM指数(K-means) ：", FMI)  # 绘制k-means结果
x0 = X[label_pred == 0]
x1 = X[label_pred == 1]
x2 = X[label_pred == 2]
 plt.scatter(x0[:, 0], x0[:, 1], c="red", marker='o', label='label_0')
 plt.scatter(x1[:, 0], x1[:, 1], c="green", marker='*', label='label_1')
 plt.scatter(x2[:, 0], x2[:, 1], c="blue", marker='+', label='label_2')
 plt.xlabel('sepal length')
 plt.ylabel('sepal width')plt.legend(loc=2)
 plt.show()

AGNES实验代码

from sklearn import datasetsfrom sklearn.cluster 
import AgglomerativeClustering
import matplotlib.pyplot as plt
from sklearn.metrics import fowlkes_mallows_score 

iris = datasets.load_iris()
irisdata = iris.data 
clustering = Agglomerative
Clustering(linkage='ward', n_clusters=3)res = clustering.fit(irisdata) # FMI的值越大，代表聚类效果越好
FMI =fowlkes_mallows_score(iris.data[:,-1],clustering.labels_)
print("FM指数(AGNES) ：", FMI)
 plt.figure()
 d0 = irisdata[clustering.labels_ == 0]
 plt.plot(d0[:, 0], d0[:, 1], 'r.')
 d1 = irisdata[clustering.labels_ == 1]
 plt.plot(d1[:, 0], d1[:, 1], 'go')
 d2 = irisdata[clustering.labels_ == 2] 
 plt.plot(d2[:, 0], d2[:, 1], 'b*')
 plt.xlabel("Sepal.Length")
 plt.ylabel("Sepal.Width")
 plt.title("AGNES Clustering")plt.show()

实验结果


DBSCAN的实验代码

import matplotlib.pyplot as plt
import numpy as np
from sklearn.cluster import KMeans
from sklearn import datasets
from sklearn.cluster import DBSCAN
from sklearn.metrics import fowlkes_mallows_score

 iris = datasets.load_iris()
 X = iris.data[:, :4]  
 # #表示我们只取特征空间中的4个维度  
 dbscan = DBSCAN(eps=0.4, min_samples=9)
 dbscan.fit(X)
 label_pred = dbscan.labels_ # FMI的值越大，代表聚类效果越好

FMI =fowlkes_mallows_score(iris.data[:,-1],label_pred)
print("FM指数(DBSCAN) ：", FMI)
 x0 = X[label_pred == 0]
 x1 = X[label_pred == 1]
 x2 = X[label_pred == 2]
 plt.scatter(x0[:, 0], x0[:, 1], c="red", marker='o', label='label0')
 plt.scatter(x1[:, 0], x1[:, 1], c="green", marker='*', label='label1')
 plt.scatter(x2[:, 0], x2[:, 1], c="blue", marker='+', label='label2')
 plt.xlabel("Sepal.Length")
 plt.ylabel("Sepal.Width")plt.title("DBSCAN")
 plt.legend(loc=2)
 plt.show()    

总结：我们使用鸢尾花数据集，通过运用三个聚类算法的比较，通过指标FMI 来评测实验结果得优劣，从实验结果看出，在鸢尾花数据集中，K-means的表现最好。