Kernelized General Fuzzy c-Means Clustering

The general algorithm to run Kernel k-Harmonic Means and Kernel Fuzzy c-Means.

Paper Source: Gupta A., Das S., "On the Unification of k-Harmonic Means and Fuzzy c-Means Clustering Problems under Kernelization", in the 2017 Ninth International Conference on Advances in Pattern Recognition (ICAPR 2017), pp. 386-391, 2017.

Example of usage

>>> import numpy as np
>>> X = np.array([
...    [10, 11], [12, 13], [14, 15], [-10, -11], [-12, -13], [-14, -15]
... ])
>>> y = np.array([0, 0, 0, 1, 1, 1])
>>> k = 2
>>> centers, mem, cost = kernel_gfcm(X, n_clusters=k)
>>> print(centers)
[[ 12.  13.]
[-12. -13.]]
>>> print(mem)
[0 0 0 1 1 1]
>>> print(cost)
1.981515079547317
>>> from sklearn.metrics import adjusted_rand_score as ARI
>>> print('ARI =', ARI(y, mem))
ARI = 1.0

Experiment on data with large number of clusters

The folder large_num_clusters contains the files to conduct the experiment (reported in the source paper) comparing the performance of Kernel k-Harmonic Means with Kernel Fuzzy c-Means, Kernel k-Means, and BIRCH.