Model Explainability

The Arize platform can help you understand why your model produced its predictions.

Sending Feature Importance

Arize supports 2 methods for ingesting and visualizing feature importance

Method

User Calculated SHAP

Method

Surrogate Model

Additional Resources

• Blog: What Are Global, Cohort and Local Model Explainability?

• Blog: Overcoming AI's Transparency Paradox

Questions? Email us at [e-mail] or Slack us in the #arize-support channel