Synthetic Minority Over-sampling Technique (SMOTE)

Version 1.0 (63,6 ko) par Bjarke Skogstad Larsen

Synthetic Minority Over-sampling Technique, DOI: https://doi.org/10.1613/jair.953

https://github.com/dkbsl/matlab_smote

Suivre

5.0

(4)

1,6K téléchargements

Mise à jour 7 mai 2020

Afficher la licence sur GitHub

SMOTE: Synthetic Minority Over-sampling Technique
This function is based on the paper referenced (DOI) below - with a few additional optional functionalities.
DOI: https://doi.org/10.1613/jair.953

This function synthesizes new observations based on existing (input) data, and a k-nearest neighbor approach. If multiple classes are given as input, only neighbors within the same class are considered.
This function can be used to over-sample minority classes in a dataset to create a more balanced dataset.

Citation pour cette source

Bjarke Skogstad Larsen (2024). Synthetic Minority Over-sampling Technique (SMOTE) (https://github.com/dkbsl/matlab_smote/releases/tag/1.0), GitHub. Récupéré le avril 19, 2024.