Synthetic Minority Oversampling Technique (SMOTE)
Definition
SMOTE supplements rare classes in training with artificial examples between similar existing cases.
SMOTE generates synthetic examples of a minority class by interpolation between neighboring training points.
Simply explained
Additional intermediate examples are generated from a rare group to increase its representation during training.
Why it matters in practice
The term helps to select, develop, evaluate AI systems in a well-founded manner and realistically assess their technical risks.
Example from business
A failure model supplements rare error cases in the training set and evaluates the result on unchanged test data.
Limits and pitfalls
- SMOTE can cause class overlap, amplify noise, and must not be applied before splitting into training and test sets.
Origin & history
Chawla and co-authors published SMOTE in 2002 as a method for imbalanced classification data.
Use in mid-sized companies
- Possible example: A failure model supplements rare error cases in the training set and evaluates the result on unchanged test data.
Frequently asked questions
What does Synthetic Minority Oversampling Technique (SMOTE) mean?
SMOTE generates synthetic examples of a minority class by interpolating between neighboring training points. What should be considered? SMOTE can cause class overlap, amplify noise, and must not be applied before splitting into training and test sets.
Sources
- SMOTE: Synthetic Minority Over-sampling Technique · Chawla et al., SMOTE: Synthetic Minority Over-sampling Technique
Source checked on 10.10.2026. Editorial application examples are hypothetical; no evidence of real customer results.
Editorially reviewed on: