Question

I am currently working on a project that requires to solve the following problem:

Let’s say that each time a user access a specific resource on the network from his mobile device, a system stores his position (latitude, longitude).

Then I need to be able to tell the user the “areas” from which he use to access that resource. And by “areas” y could mean circumferences (center & ratio).

The problem is I need some criteria to discard isolated points and I need some criteria for grouping remaining points into “areas”.

I bet there must be some literature on the subject, but the issue is I dont even know the terms to refer this problem, and where to start looking.

Thanks in advance 😉

Answer 1

You might be interested in clustering and/or outlier detection algorithms. The simplest option might be the DBSCAN algorithm, which finds clusters in a dataset while ignoring points that are not sufficiently “connected” to other points.