In the SMOTE paper, the authors present the logic of creating synthetic examples when all features are nominal (section 6.2, SMOTE-N):

To generate new minority class feature vectors, we can create new set feature values by taking the majority vote of the feature vector in consideration and its k nearest neighbors

Along with this example:

Let F1 = A B C D E be the feature vector under consideration and let its 2 nearest neighbors be

F2 = A F C G N

F3 = H B C D N

The application of SMOTE-N would create the following feature vector: FS = A B C D N

How would FS be chosen in the case that F3 = H B C I N? How does Value Difference Metric by Cost and Salzberg described in the paper assist in this case?


I don't know what "VDM" stands for, but a simple solution for tie breaking is to randomly pick one of the tied options.

  • $\begingroup$ Edited my question - VDM: Value Difference Metric by Cost and Salzberg described in the paper. See the paper by C&S: cs.utsa.edu/~bylander/cs6243/cost93weighted.pdf $\endgroup$
    – shakedzy
    Jan 7 '18 at 22:44
  • $\begingroup$ I guess I'll have to email the authors of the paper.. in the meantime, your answer seems legit :) $\endgroup$
    – shakedzy
    Jan 7 '18 at 22:46

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.