facet#
Human-explainable AI.
This is the class and function reference of FACET for advanced model selection, inspection, and simulation.
The figure below provides a high level overview of the workflow when using FACET, and for each step in the workflow, a brief description.
Please refer to the tutorials for examples of using FACET classes and functions, and the release notes for recent API updates and bug fixes.
Submodules#
Basic data management for FACET's enhanced machine learning workflow. |
|
Factory classes for common SHAP explainers. |
|
Learner inspection with standard local SHAP and global FACET metrics describing how features combine to contribute to all model predictions. |
|
Learner selection with hyperparameter optimization. |
|
Historical univariate simulation of changes to predicted outcome(s) based on selected fixed values of a selected input feature. |
|
Bootstrap cross-validation, including a stationary version for use with time series data. |