RafeyIqbalRahman / Clustering-Techniques

This repository demonstrates several clustering techniques using Scikit-Learn.

Clustering Techniques

Clustering refers to the act of placing data points with similar characteristics into groups. It can be simply referred to as the classification of data points. Although there are several clustering techniques, in this repository, 10 of such techniques have been implemented on a fake, generated data.

Objective

The objective of the implementation was to test these techniques on fake data and see which techniques perform perfect clustering (creating no more or no less than 2 clusters in this case).

The clustering techniques implemented are as follows along with the respective clusters:

1. AffinityPropagation. Clusters: [0, 0, 0, 1, 1, 1]
2. AgglomerativeClustering. Clusters: [1, 1, 1, 0, 0, 0]
3. Birch. Clusters: [1, 1, 1, 0, 0, 0]
4. DBSCAN. Clusters: [-1, -1, -1, -1, -1, -1]
5. GaussianMixture. Clusters: [0, 0, 0, 1, 1, 1]
6. KMeans. Clusters: [1, 1, 1, 0, 0, 0]
7. MeanShift. Clusters: [2, 0, 1, 3, 4, 5]
8. MiniBatchKMeans. Clusters: [0, 0, 0, 1, 1, 1]
9. OPTICS. Clusters [0, 0, 0, -1, -1, -1]
10. SpectralClustering. Clusters: [1, 1, 0, 0, 0, 0]