A method that provides sets of if-then rules locally sufficient for a model to make a prediction with high confidence, including precision, coverage, and complexity metrics.

A visualization showing the global impact of each feature on a model's predictions, calculated using an approximation of Shapley values.