Toward the Development of the Sand Dune Clustering Algorithm

This paper proposed a new partition-based clustering algorithm inspired by the sand dunes. The proposed algorithm is called the Sand Dune Clustering Algorithm. The algorithm accepts an r x c matrix of data where one of its columns should be identified as the weight. The weight of the column is then utilized to cluster the data points. Based on initial experimentation, the algorithm was able to group the sample points. Furthermore, it was disclosed that the algorithm is relatively fast, running at the order O(rc + r). Thus, it can be concluded that the proposed algorithm is a promising clustering algorithm. Limitations and future work are also discussed.