@validbeck thank you for reviewing and pushing some changes! Much appreciated.

I propose two new articles under Ongoing monitoring

Agreed, it makes sense to split this topic now. I was originally trying to create a placeholder to iterate over in the future, but there's too much content already. Splitting will address this issue.

... page is already so text-heavy ...

Related, I had some Mermaid diagrams in there at some point but removed them to save some time. I might see if I can include these to reduce the feeling of text heaviness a bit.

Looks great :) A couple of minor suggestions:

• Use Ongoing Monitoring in the glossary instead of Monitoring
• To aligned ourselves a bit with SR11-7, replace effectiveness by robustness

For example in here:

I suggest:

Ongoing Monitoring
To regularly evaluate the ongoing accuracy, robustness, and stability of a model after it has been deployed.

Ongoing Monitoring
Monitoring of model performance is a critical component of model risk management, as emphasized in SR 11-7. This process involves regularly assessing a model’s predictive accuracy, robustness, and stability to ensure that it continues to perform as expected and remains fit for its intended purpose after deployment.

Same here:

I suggest the same text or similar:
Monitoring of model performance is a critical component of model risk management, as emphasized in SR 11-7. This process involves regularly assessing a model’s predictive accuracy, robustness, and stability to ensure that it continues to perform as expected and remains fit for its intended purpose after deployment.

Looks great :) A couple of minor suggestions:

• Use Ongoing Monitoring in the glossary instead of Monitoring
• To aligned ourselves a bit with SR11-7, replace effectiveness by robustness

For example in here:

I suggest:

Ongoing Monitoring To regularly evaluate the ongoing accuracy, robustness, and stability of a model after it has been deployed.

Ongoing Monitoring Monitoring of model performance is a critical component of model risk management, as emphasized in SR 11-7. This process involves regularly assessing a model’s predictive accuracy, robustness, and stability to ensure that it continues to perform as expected and remains fit for its intended purpose after deployment.

Same here:

I suggest the same text or similar: Monitoring of model performance is a critical component of model risk management, as emphasized in SR 11-7. This process involves regularly assessing a model’s predictive accuracy, robustness, and stability to ensure that it continues to perform as expected and remains fit for its intended purpose after deployment.

We might not necessarily want to stick to only talk about SR 11-7 since other regulations such as SS 1/23 and E23 are relevant and applicable to our customers. Maybe there's a way to make these statements more generic or somehow say things like "as emphasized in regulations such as X, Y, Z"?

Looks great :) A couple of minor suggestions:

• Use Ongoing Monitoring in the glossary instead of Monitoring
• To aligned ourselves a bit with SR11-7, replace effectiveness by robustness

For example in here:

I suggest:
Ongoing Monitoring To regularly evaluate the ongoing accuracy, robustness, and stability of a model after it has been deployed.
Ongoing Monitoring Monitoring of model performance is a critical component of model risk management, as emphasized in SR 11-7. This process involves regularly assessing a model’s predictive accuracy, robustness, and stability to ensure that it continues to perform as expected and remains fit for its intended purpose after deployment.
Same here:

I suggest the same text or similar: Monitoring of model performance is a critical component of model risk management, as emphasized in SR 11-7. This process involves regularly assessing a model’s predictive accuracy, robustness, and stability to ensure that it continues to perform as expected and remains fit for its intended purpose after deployment.

We might not necessarily want to stick to only talk about SR 11-7 since other regulations such as SS 1/23 and E23 are relevant and applicable to our customers. Maybe there's a way to make these statements more generic or somehow say things like "as emphasized in regulations such as X, Y, Z"?

I agree with Andres' comments. I'm happy to go with either a generic approach or include references to specific MRM frameworks. Since all the regulations emphasize monitoring, mentioning SS1/23, SR 11-7, TRIM, and E-23 should cover the key guidelines.

Updates to address review feedback

@validbeck if these updates pass muster with you, could you approve this PR, please?

Updated wording

Now refers to "ongoing monitoring" in the glossary, uses "robustness" instead of "effectiveness", and includes references to sample regulations:

Separate task topics

Reworked overview topic with some sample output from testing and child topics included below:

Separate task topic for enabling monitoring:

Separate task topic for reviewing monitoring results:

PR Summary

This pull request introduces significant enhancements to the documentation by adding comprehensive sections on monitoring capabilities. The key changes include:

1. New Monitoring Section: A new section titled 'Monitoring' has been added to the documentation, which includes detailed guides on enabling and reviewing ongoing monitoring for models.

   • Enable Monitoring: Instructions on how to add monitoring=True to the code snippet and select a monitoring template.
   • Review Monitoring Results: Steps to access and review the monitoring results within the ValidMind Platform UI.

2. Glossary Updates: New terms related to monitoring have been added to the glossary, including definitions for backtesting, model drift, model performance, ongoing monitoring, recalibrating models, and reporting and governance.

3. Sample Monitoring Results: Images and examples of monitoring results have been included to provide a visual understanding of the monitoring process.

4. Test Descriptions: The test descriptions section has been updated to include ongoing monitoring tests, such as Feature Drift, Prediction Across Each Feature, Prediction Correlation, and Target Prediction Distribution Plot.

5. Code Samples: The code samples have been updated to include examples of how to enable and use monitoring in the ValidMind platform.

6. Miscellaneous Updates: Minor updates to existing documentation to ensure consistency and clarity, including changes to the glossary and various sections to reflect the new monitoring capabilities.

Test Suggestions

• Verify that the new 'Monitoring' section is accessible from the main documentation page.
• Test the code snippets provided in the 'Enable Monitoring' section to ensure they work as expected.
• Check the links in the 'Review Monitoring Results' section to ensure they navigate to the correct pages.
• Validate the glossary entries for new monitoring terms to ensure they are accurate and comprehensive.
• Run the updated code samples to confirm that monitoring can be enabled and results can be reviewed as described.
• Ensure that the images and examples of monitoring results are correctly displayed and provide clear insights.

@validbeck I updated this PR with the new style for the "Prerequisites" section, FYI.