2001, Knowledge Discovery, Presentation and Testing

ML internals: Synthetic Minority Oversampling (SMOTE) Technique

Domino Data Lab

MAY 20, 2021

We present the inner workings of the SMOTE algorithm and show a simple “from scratch” implementation of SMOTE. Their tests are performed using C4.5-generated 2002) do not present a rigorous mathematical treatment for this modification, and the suggested median correction appears to be purely empirical-driven.

Machine Learning

Machine Learning Metrics Data mining Knowledge Discovery

Data Leaders Brief

ML internals: Synthetic Minority Oversampling (SMOTE) Technique

Webinars

Stay Connected