The algorithm actually is saying arbitrarily select upon a p

that means at the very beginning you can select at any point starting point.

Then you retrieve all the points density-reachable from p

with respect to Epsom and the minimum number of points.

That means you try to find from this p how far it can density-reachable.

That means you first find directly density-reachable

then find its transitive closure or the reachable y, okay?

That means you find a core and then you find the border.

If p is a core point then the cluster is formed because it

will reach from anywhere, you'll find you can't reach it.

But if p is a border point, like this one is a border point, this one is a border

point, then no point at density- reachable from p for example, like this point, okay?

No point it can't be density-reachable, then you were think to this one is,

noise [INAUDIBLE].

Then the DBSCAN where the, the,

the next point in the database, it means you try to find a new point to start.

You will continue this process until all the points

have been processed so this is the output.