K-Means Algorithm

K-Means algorithm is one of the most-commonly used clustering algorithms. Clustering algorithms try to group similar data points (may have various meainings) with respect to a selected criteria. These algorithms are widely used in data mining, pattern recognition, image analysis, supply chain management, etc. Like all clustering algorithms, K-Means algorithm tries to assign data points into a determined number of clusters, K. The algorithm assumes that each cluster has a center (centroid) and aims to minimize the sum of distances of data points from the centroids of clusters that the data points are assigned to. This algorithm is a fast algorithm that converges rapidly with good solutions. Random initialization prevents the algorithm from being stuck to a bad solution. This template runs the algorithm with VBA code, therefore macros need to be enabled. The algorithm may be run either once at a time or each step may be observed in order to observe the convergence process of the alg.

 http://spreadsheetzone.com/288/KMeans-Algorithm.html