This paper suggests that one can use principal components analysis to assess the local dimensionality in a complex dataset. If this is large, then data mining is unlikely to be successful. If it is small, then various statistical techniques may be able to uncover hidden structure.
American Statistical Association Section on Computational Statistics
curse of dimensionality, data mining, nonparametrics
and Olszewski, R.
Estimating Local Dimensionality, American Statistical Association Section on Computational Statistics, [online], https://tsapps.nist.gov/publication/get_pdf.cfm?pub_id=151731
(Accessed December 1, 2023)