Intermediate · Data
Downsampling
Visual diagram · (in preparation) · Math · (in preparation) · Worked example · 3 difficulty levels.
TL;DR. Reducing the number of data points, either by lowering resolution or selecting fewer examples from an over-represented class.
Technical Definition
Reducing the number of data points, either by lowering resolution or selecting fewer examples from an over-represented class.
How it works
Downsampling can refer to two distinct processes: reducing the dimensionality or resolution of data (like shrinking image size) or decreasing the number of samples in an over-represented class within a dataset. In imbalanced datasets, downsampling the majority class helps prevent the model from being biased towards it and improves performance on the minority class.
Related Concepts
- Feature Selection — Choosing the most useful subset of features to improve performance and interpretability.
- Data Imbalance — When some classes or groups are vastly more frequent in the dataset than others.