A modified self-updating clustering algorithm for application to dengue gene expression data

This work proposed a conceptually simple and computationally straightforward clustering algorithm based on the Cauchy-type distance for data clustering. It was demonstrated that the proposed approach does not require the priori number of clusters and the convergence of the proposed algorithm was proved. The experiment results showed that the proposed clustering algorithm was superior to other compared algorithms. Computational complexity was also provided. A real dengue gene expression dataset was used to demonstrate the effectiveness of the proposed method.