Analytic Transparency: Creating Trust with Explainability and Model Monitoring
Predictive models are increasingly used to make decisions that directly affect humans. Naturally, concerns about trust and reliability arise.
In this session, we address these issues by demonstrating tools that we have developed for model-explainability and model-monitoring. We also show how our solutions have been implemented as functions in the new DSML Toolkit for Python that can be invoked from any Python environment.