Detecting hidden structures in a dataset with clustering

A large part of unsupervised learning is devoted to the clustering problem. The goal is to group similar points together in a totally unsupervised way. Clustering is a hard problem, as the very definition of clusters (or groups) is not necessarily well posed. In most datasets, stating that two points should belong to the same cluster may be context-dependent or even subjective.

There are many clustering algorithms. We will see a few of them in this recipe, applied to a toy example.

How to do it...

1. Let's import the libraries:

   In [1]: from itertools import permutations
           import numpy as np
           import sklearn
           import sklearn.decomposition as dec
           import sklearn.cluster as clu
           import sklearn.datasets as ds
           import sklearn.grid_search as gs
           import matplotlib.pyplot as plt
           %matplotlib inline

2. Let's generate a random dataset with three clusters:

   In [2]: X, y = ds.make_blobs(n_samples=200, n_features...