ML internals: Synthetic Minority Oversampling (SMOTE) Technique
Domino Data Lab
MAY 20, 2021
Machine Learning algorithms often need to handle highly-imbalanced datasets. This renders measures like classification accuracy meaningless. This in turns makes the performance evaluation of the classifier difficult, and can also harm the learning of an algorithm that strives to maximise accuracy. Machine Learning, 57–78.
Let's personalize your content