SMOTE

SMOTE: Synthetic Minority Over-sampling Technique.

smote(x, y, k = 5L, ndups = 0L)

Arguments

x

matrix, features

y

factor, class (length has to be equal to number of rows in x).

k

numeric(1), number of nearest neighbours.

ndups

numeric(1), number of synthetic duplicates to upsample minority. Default is 0 which tries to upsample to majority.

Value

list with elements x features with SMOTEd samples and y with added minority class elements.

References

Chawla, Bowyer, Hall and Kegelmeyer. 2002. SMOTE: Synthetic Minority Over-sampling Technique. Journal of Artificial Intelligence Research 16. 321–357. doi: 10.1613/jair.953