SMOTE Save

This is a README file

The code is an implementation of the SMOTE model(Synthetic Minority Over-sampling Technique) from the paper N. V. Chawla, K. W. Bowyer, L. O.Hall, W. P. Kegelmeyer, “SMOTE: synthetic minority over-sampling technique,” Journal of artificial intelligence research, 321-357, 2002.

Usage

N = percentage of over-sampling required k = no. of nearest neighbors smote_test = Smote('euclidian') smote_test.genarate_synthetic_points(min_samples,N,k)

Note that ball tree uses an implementation of sklearns nearest neighbor module.In case you do not hav sklearns nearest neighbor module you can implement the euclidian distance to find the nearest neighbor.

Requirements: NumPy