Working with artifacts

Use ValidMind Platform to log thorough artifacts as you validate your models. From status and severity to proposed remediation plans and due dates, ​ValidMind allows you to oversee the minutiae to ensure organizational compliance.

Key concepts

Model artifacts (findings) are detailed observations identified during the model validation process, highlighting any major or minor issues, deficiencies, model limitations, stability and robustness concerns, or needed adjustments.

• These artifacts are critical for understanding the risk exposure and compliance status of models within an organization.
• To make them easier to track, artifacts are typically categorized by risk area, business unit, model status, and individual model, enabling targeted resolution and informed decision-making to mitigate identified risks and ensure model reliability and accuracy.

Good artifacts help explain how well the model performs in terms of predictive accuracy, generalization to new data, and the importance of variables or features used in the modeling process, and typically include:

performance metrics

Quantitative measures that evaluate how well the model predicts outcomes, such as accuracy, precision, and recall.

feature importance

The ranking of input variables based on their contribution to the predictive power of the model.

model interpretability

Insights into how the model makes its predictions, which can help in understanding the decision-making process.

validation results

Artifacts from testing the model on unseen data, for example cross-validation results.

residual analysis

Investigating the errors or discrepancies between predicted and actual outcomes.

Manage artifacts

What’s next

• Preparing validation reports