Improve Basic Sequential Algorithm Scheme Using Ant Colony Algorithm

Basic Sequential Algorithm Scheme BSAS is a sequential algorithm for data clustering. It is suitable for unraveling compact dataset. The BSAS algorithm is sensitive to the order of data presentation; different clustering results could be produced if the input data are presented in a different order. Because the number of clusters in the results varies depending on the value of threshold, multiple run is one of the solutions to obtain optimal threshold. In this paper, BSAS is optimized using Ant Colony Optimization ACO Algorithm to solve the order sensitivity problem. The new proposed algorithm obtains the best order from ACO algorithm, which is based on the calculations of minimum distances between points, and passes the optimal order to BSAS algorithm as an input order. Finally, the proposed algorithm is compared and verified using the Sum Square Error SSE. The experimental results show that the proposed algorithm developed the BSAS algorithm.