« Back to Glossary Index

Explainability is how to take a ML model and explain what happens inside it from input to output. For examples, using model agnostic methods, such as SHAP or LIME, one can discover meaning between input data attributions and model outputs: this is usually done by taking a single input value of predictions, perturbing it, and understanding how perturbations in a model’s inputs affect the end-prediction of the model.

See Interpretable Machine Learning for further details.