Model Validation

Do you have a vendor or internally developed model that you're evaluating, testing, or have already put into production?  Here are some signs that you're due for a model validation:

  • You have never had a model validation performed

  • Your board or regulator wants to ensure that the risks associated with your predictive model are accounted for from both a quantitative and qualitative perspective

  • You have received a vendor model that you feel may be statistically weak

  • You have received a vendor model that is statistically robust but is lacking detailed, interpretable documentation

  • Your model isn't making predictions with the accuracy you'd like it to, or you are concerned you have not been comparing model predictions to actual results proficiently

  • You have continued to collect data in your organization, but the model you are using has not changed or leveraged this new data to improve accuracy

If any of these scenarios sound like your organization, let us help you unlock the true return on your predictive modeling investment.

LET'S TALK MODEL VALIDATION

© 2019 by Ketchbrook Analytics

  • Black LinkedIn Icon