A grid clustering algorithm using cluster boundaries

Grid-based clustering methods have been extensively applied on large data sets because of low computational cost. In this paper, we propose a new algorithm for grid-based clustering by finding the optimal grid-size using the boundaries of the clusters. The algorithm has linear time complexity. The problem of outliers is resolved with the help of local outlier factor (LOF). We apply the proposed method on various synthetic as well as biological data sets. The results are compared with K-means and few existing grid-based techniques. The comparison results show the effectiveness of the proposed method.