If I want to use an undersampling approach to construct the machine learning model (classification model predicting whether the target compounds are active or inactive, using chemical fingerprint as variables), I am wondering if there are any criteria to determine how many times I should sample the data from the majority group (the minority is 14% and the majority is 86%) and build the ML model? I am working with biological assay data and I am recommended not to use oversampling approaches. To determine which sampling approach we should use, are there any criteria to determine before constructing the ML model? Is it really dependent on the field of study?

  • $\begingroup$ My similar question has been moved here, stats.stackexchange.com/questions/540770/… $\endgroup$
    – tassaneel
    Commented Aug 20, 2021 at 2:05
  • $\begingroup$ Another option is Matter Modeling Stack exchange. This question isn’t really about chemistry but modeling so it typically isn’t appropriate. $\endgroup$
    – Cody Aldaz
    Commented Sep 2, 2021 at 17:35


Your Answer

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