After model deployment, to continuously improve the model based on real-world performance data.

Before training the model, to optimize the model's performance and ensure it meets regulatory standards from the outset.

During data collection, to adjust the model's parameters based on the incoming data quality and volume.

After model evaluation, to refine the model based on its performance metrics and compliance checks.

Both before training the model and after model evaluation, to initially optimize and then further refine the model's performance and compliance.