Alternative Fuzzy C-Means Algorithms

I have successfully tackled an optimization challenge by developing an extended version of the Fuzzy C-means algorithm, known as Alternative Fuzzy C-means. I drew inspiration from the research published in the Journal of Pattern Recognition Society in Machine Learning for this project.

In this endeavor, I introduced a novel matrix for centroid-to-data point distance computation, a key element that significantly enhances the accuracy of the model. Notably, I achieved remarkable results by building the entire model from scratch, bypassing the use of external libraries like scikit or Tensor. This achievement led to a substantial accuracy improvement, elevating it from 40% to an impressive 80%.

New matrix has been introduced to calculate the distance between centroids and data points.

Please feel free to add your logic in-order to increase the accuracy