1

For my theoretical work, I have generated synthetic data where a single cluster is completely encircled by other clusters, in 2D the data set has the form (roughly) as in the picture below:

data

I am looking for some real data that have a similar characteristic, dimensionality is not of importance (for visualization I can use PCA and project to 3D).

I have looked through most of the data sets in UCI and randomly searched on the internet but could not find what I am looking for...

Intuitively I would expect that many rich data sets should have this layout, that is one cluster of similar objects and all the rest of the data which is a noise in which the cluster is immersed.

I would appreciate it if someone could point me out which data usually has this layout (medical data, econometrics, imaging, geospatial data) any other suggestions would be very appreciated!

0

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.