Mastering Machine Learning Algorithms

Example of fuzzy C-means with Scikit-Fuzzy

Scikit-Fuzzy (http://pythonhosted.org/scikit-fuzzy/) is a Python package based on SciPy that allows implementing all the most important fuzzy logic algorithms (including fuzzy C-means). In this example, we continue using the MNIST dataset, but with a major focus on fuzzy partitioning. To perform the clustering, Scikit-Fuzzy implements the cmeans method (in the skfuzzy.cluster package) which requires a few mandatory parameters: data, which must be an array D ∈ ℜ^{N × M} (N is the number of features; therefore, the array used with Scikit-Learn must be transposed); c, the number of clusters; the coefficient m, error, which is the maximum tolerance; and maxiter, which is the maximum number of iterations. ...