Less is more approach for balanced minimum sum-of-squares clustering
Balanced clustering; minimum sum-of-squares; optimization
After the advance in collecting and storing data and the increase on applications which are source of new information, the number of data elements available is very huge either on volume as on variety. Because the rising in the data quantity, the necessity of understand and summarize them becomes fundamental. Given a set of points, balanced Minimum sum-of-squares Clustering(balanced MSSC) aims to find subsets, denoted clusters, miniminzig the summation of squared distances from each data point to the centroid of its cluster. In this work, we present a variable neighborhood search heuristic for balanced MSSC following the recently proposed LIMA (less is more approach). Computational experiments show that the proposed heuristic outperforms the current state-of-the-art algorithm for the problem.