Input: sample data set DS, parameters (Eps, Minpts) Output: data cluster set C DBSCAN (DS, Eps, Minpts) Begin Mark all points in DS as unvisited; Do Randomly choose an unvisited xi; Mark xi as visited; If points in xi’s $\mathrm{Eps}$-neighborhood are no less than Minpts  Create a new data cluster (DC);  Set N consist of points in xi’s $\mathrm{Eps}$-neighborhood;   For each point xj in N    If xj is unvisited     Mark xj as visited;     If points in xj’s Eps-neighborhood are no less than Minpts, add points to N;     End if     If xj is not a member of any data cluster, add xj to DC;     End if    End if   End for;   Output DC; Else mark xi as a noise point; End if Until all unvisited points are visited; Output C End