What method to use for cluster identification ?

Question

This question is from a confused novice.

I have a data set with where each point is located in a 2-D space defined by two objectives (say, X and Y). I wish to identify a set of points from this space such that their distance from X AND y is minimum.

What can be the most efficient / simple / targeted approach to achieve this ? I started with k-means clustering but eventually ventured into material related to multiobjective optimization, multiobjective clustering, clustering ensemble, etc.

K-means CA will try to identify clusters via unspecified leaning of the data based on internal criteria (compactness and separation), but is there a way to identify clusters using external criteria such as distance of a cluster member from the above X and Y axes?

Will the multiobjective methods be an overkill ? Are the advanced methods unobtrusive? Can these methods work with the dataset that already exists ?

A lead in the right direction will be greatly appreciated. Thank you.

Answer 1

It does not sound as if you are looking for clusters.

If you want to find objects close to (0,0), this is a straightforward data ordering problem. Think of it as sorting.

distance from X AND y is minimum

Minimizing the distance from axes is not a clustering objective. Use optimization methods. Define your objective mathematically.