global explanation

Terms from Artificial Intelligence: humans at the heart of algorithms

The glossary is being gradually proof checked, but may have typos and misspellings.

A global explanation is a form of explainable AI, which attempts to explain the behaviour of a model on any input. For example, some variants of SHAP assess the importance each input feature on average over all input–output pairs.. This is in contrast to a local explanation, which attempts to explain a particular input–output of the model.

Used in glossary entries: explainable AI, local explanation, SHAP