A- example of unsupervised learning algorithms: hierarchical clustering

The basic principles of hierarchical clustering

 

⼀ kinds of small samples for accurate clustering clustering algorithms: hierarchical clustering. The so-called more and more precise, ⼀ aspect refers to the clustering method full transparency in the implementation process, the other side is ⼀ in a real application scenarios for the amount of data (about thousands of rows) of data, hierarchical clustering to have a clustering effect of looking great!

 

First of all it will calculate the nearest data point, be classified as a cluster, and then take the center of mass of the cluster as a representative to participate in the next time the process of selecting the nearest point, in short, it is the constant pairwise merge until it merged into a cluster so far

 

Fake code

1: Calculation proximity twenty-two element

2： repeat

3: merge the two closest clusters

4: more proximity matrix updated to reflect the proximity between the original and the new cluster or cluster elements

5: Until leaving only a cluster ⼀

In [2]:

import numpy as np
import pandas as pd
import matplotlib.pyplot as plt
%matplotlib inline

 

In [4]:

from sklearn.datasets import load_iris
iris = load_iris()

In [6]:

iris.data

In [7]:

iris.target

Out[7]:

array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
       0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
       0, 0, 0, 0, 0, 0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
       1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
       1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2,
       2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2,
       2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2])

In [8]:

from sklearn.cluster import AgglomerativeClustering
agClustering = AgglomerativeClustering(n_clusters=3)

In [9]:

agClustering.fit(iris.data)

Out[9]:

AgglomerativeClustering(affinity='euclidean', compute_full_tree='auto',
            connectivity=None, linkage='ward', memory=None, n_clusters=3,
            pooling_func=<function mean at 0x0000027004F287B8>)

In [12]:

agClustering.labels_

Out[12]:

array([1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
       1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
       1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
       0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 2, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
       0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 2, 0, 2, 2, 2, 2, 0, 2, 2, 2,
       2, 2, 2, 0, 0, 2, 2, 2, 2, 0, 2, 0, 2, 0, 2, 2, 0, 0, 2, 2, 2, 2,
       2, 0, 0, 2, 2, 2, 0, 2, 2, 2, 0, 2, 2, 2, 0, 2, 2, 0], dtype=int64)

In [15]:

from sklearn.metrics import accuracy_score  

In [16]:

accuracy_score(iris.target,agClustering.labels_)

Out[16]:

0.23333333333333334