alexein777/K-MESA

Advanced K-Means clustering algorithm using soft computing and regularization ideas

K-MESA (K-Means with Simulated Annealing) is an advanced clustering algorithm that uses simulated annealing to help algorithm avoid local optimums in updating cluster centroids. Annealing is applied in centroid update step and algorithm convergence is observed across different (mostly simple) datasets. Results are compared with simple K-algorithm and decent conclusion is inferred.