天津理工大学辨识技术实验四 + 动态聚类（附源代码）

import numpy as np import matplotlib.pyplot as plt import pandas as pd from sklearn.datasets import load_iris from sklearn.cluster import KMeans from sklearn.metrics import silhouette_score iris_dataset = load_iris() #内置函数载入iris数据集 iris_fs = iris_dataset['data'] #取出iris数据集的特征 iris_lb = iris_dataset['target'] #取出iris数据集的标签 x1 = iris_dataset.data[:, 0] # X- Axis -sepal length y1 = iris_dataset.data[:, 1] # Y- Axis - sepal width x2 = iris_dataset.data[:, 2] # X- Axis - petal length y2 = iris_dataset.data[:, 3] # Y- Axis - petal width species = iris_dataset.target # Species 标签数据 # Scatterplot 对原始数据进行可视化，分别以花萼特征(长、宽)和花瓣特征（长、宽）进行二维绘图 plt.figure(figsize=(10,8)) plt.subplot(221) plt.title("Iris Dataset - Classification By Sepal Sizes") plt.scatter(x1, y1, c = species) plt.xlabel('Sepal length') plt.ylabel('Sepal width') plt.xticks(()) plt.yticks(()) # 根据Sepal（花萼）绘制散点图，观察原始信息 plt.subplot(222) plt.title("Iris Dataset - Classification By Petal Sizes") plt.scatter(x2, y2, c = species) plt.xlabel('Petal length') plt.ylabel('Petal width') plt.xticks(()) plt.yticks(()) # 根据Petal（花瓣）绘制散点图，观察原始信息 # 通过k-means进行数据的聚类 kmeans = KMeans(n_clusters = 3, init = 'k-means++', algorithm = 'auto' , max_iter = 100, n_init = 10, random_state = 2022) #关键参数n_cluster，类别中心数 y_kmeans = kmeans.fit_predict(iris_fs) #计算准确度 c = 0 for i in range(150): if(y_kmeans[i] == iris_lb[i]): c = c + 1 print("准确度：%.2f%%\n" %(c*100/150)) #计算轮廓系数，取值范围为[-1, 1]，-1表示分类效果差，1表示分类效果好 print("轮廓系数：") print(silhouette_score(iris_fs, iris_lb)) # 聚类结果用第1,2列（花萼）的特征表示 plt.subplot(223) plt.scatter(iris_fs[y_kmeans == 0, 0], iris_fs[y_kmeans == 0, 1], c="red", marker='o', label = 'Iris-setosa') plt.scatter(iris_fs[y_kmeans == 1, 0], iris_fs[y_kmeans == 1, 1], c="green", marker='o', label = 'Iris-versicolour') plt.scatter(iris_fs[y_kmeans == 2, 0], iris_fs[y_kmeans == 2, 1], c="blue", marker='o', label = 'Iris-virginica') # 聚类结果可视化简化coding plt.scatter(kmeans.cluster_centers_[:, 0], kmeans.cluster_centers_[:,1], c = 'yellow', label = 'Center') # 聚类中心 plt.title("K-means clustering plot by Sepal") plt.legend() # 聚类结果用第3,4列（花瓣）的特征表示 plt.subplot(224) fs0 = iris_fs[y_kmeans == 0] fs1 = iris_fs[y_kmeans == 1] fs2 = iris_fs[y_kmeans == 2] plt.scatter(fs0[:, 2], fs0[:, 3], c="red", marker='o', label='Iris-setosa') plt.scatter(fs1[:, 2], fs1[:, 3], c="green", marker='o', label='Iris-versicolour') plt.scatter(fs2[:, 2], fs2[:, 3], c="blue", marker='o', label='Iris-virginica') # 聚类结果可视化细化coding plt.scatter(kmeans.cluster_centers_[:, 2], kmeans.cluster_centers_[:,3], c = 'yellow', label = 'Center') # 聚类中心 plt.title("K-means clustering plot by Petal") plt.legend() plt.show()