Glossary term
Glossary term
Foundations
A type of regularization that penalizes the total number of nonzero weights in a model. For example, a model having 11 nonzero weights would be penalized more than a similar model having 10 nonzero weights.
L0 regularization is sometimes called L0-norm regularization.
Click the icon for additional notes.
Created for this library
A research team uses an L0-regularization surrogate to encourage genuine sparsity in a feature selector for tabular data.
A medical research team uses an L0-style regularizer to keep its prediction model dependent on only a small set of clinically interpretable features.
A pricing team explores L0-regularization to build a parsimonious model with the fewest features that still meets accuracy targets.
Definition source: Google for Developers Machine Learning Glossary | Creative Commons Attribution 4.0 License