Simple and crude understanding and implementation of machine learning clustering algorithm (B): clustering algorithm api initial use case

Clustering Algorithm

learning target

• Master clustering algorithm implementation process
• We know K-means algorithm theory
• We know evaluation model clustering algorithm
• The advantages and disadvantages of K-means
• Understand way clustering algorithm optimization
• Application Kmeans achieve clustering task
  

6.2 clustering algorithm api initial use

1 api Introduction

• sklearn.cluster.KMeans(n_clusters=8)
  • parameter:
    • n_clusters: the number of cluster centers began
      • Integer, default = 8, the resulting number of clusters, i.e., the center of mass produced (centroids of) number.
  • method:
    • estimator.fit(x)
    • estimator.predict(x)
    • estimator.fit_predict(x)
      • Computing cluster centers and the prediction of each sample belongs to which category, which is equivalent to calling fit (x), and then call predict (x)

2 Case

Creating different random two-dimensional data set as a training set, combined with its k-means clustering algorithm, you can try different numbers of clusters are clusters, and observe the clustering effect:

N_cluster different clustering parameters by value, different clustering results obtained

2.1 Process Analysis

2.2 code implementation

1. Create a data set

import matplotlib.pyplot as plt
from sklearn.datasets.samples_generator import make_blobs
from sklearn.cluster import KMeans
from sklearn.metrics import calinski_harabaz_score

# 创建数据集
# X为样本特征，Y为样本簇类别， 共1000个样本，每个样本4个特征，共4个簇，
# 簇中心在[-1,-1], [0,0],[1,1], [2,2]， 簇方差分别为[0.4, 0.2, 0.2, 0.2]
X, y = make_blobs(n_samples=1000, n_features=2, centers=[[-1, -1], [0, 0], [1, 1], [2, 2]],
                  cluster_std=[0.4, 0.2, 0.2, 0.2],
                  random_state=9)

# 数据集可视化
plt.scatter(X[:, 0], X[:, 1], marker='o')
plt.show()

2. using k-means clustering, and the evaluation method using CH

y_pred = KMeans(n_clusters=2, random_state=9).fit_predict(X)
# 分别尝试n_cluses=2\3\4,然后查看聚类效果
plt.scatter(X[:, 0], X[:, 1], c=y_pred)
plt.show()

# 用Calinski-Harabasz Index评估的聚类分数
print(calinski_harabaz_score(X, y_pred))