A density-grid-based method for clustering k-dimensional data