smote function

SMOTE

Synthetic Minority Oversampling Technique (SMOTE) algorithm for imbalanced classification data.

smote(y, x, k = 5, over = NULL, yminor = NULL)

Arguments

• y: Vector of response outcome as a factor

• x: Matrix of predictors

• k: Range of KNN to consider for generation of new data

• over: Amount of oversampling of the minority class. If set to NULL

  then all classes will be oversampled up to the number of samples in the majority class.

• yminor: Optional character value specifying the level in y which is to be oversampled. If NULL, this is set automatically to the class with the smallest sample size.

Returns

List containing extended matrix x of synthesised data and extended response vector y

References

Chawla, N. V., Bowyer, K. W., Hall, L. O., and Kegelmeyer, W. P. (2002). Smote: Synthetic minority over-sampling technique. Journal of Artificial Intelligence Research, 16:321-357.