0
$\begingroup$

I have developed an inventory optimization model for my warehouse and I want to know, how to validate this model, perform User acceptance testing and deploy it to production environment?

$\endgroup$
1
  • $\begingroup$ Welcome to OR.SE. It would be also great if you could show your model. It gives you more chances to view and get feedback to make your model tighter. $\endgroup$
    – A.Omidi
    Commented Feb 10 at 5:59

1 Answer 1

1
$\begingroup$

To validate your model, you need to make sure it captures the dynamics of your system correctly. To do so, you can constrain the model with your historical data (e.g. one month, or one year depending on the nature of your model), and check that it outputs what actually happened in the "real world". You can check aggregated values of variables (such as monthly flows), KPIs (costs), and more detailed values (inventory level of given product at given hour). Determining how close the details have to be to reality is an empirical question. Aiming for 100% accuracy is neither necessary nor a good idea, as you do not want to replicate noise, data errors, etc.

This way you have a baseline or a reference model, on top of which you can safely optimize by removing baseline constraints, or perhaps adding new variables to expand the search space.

This will help validate that your model is indeed correct, and make comparisons between optimizations sensible.

For user acceptance, I think you need to show them that whatever optimization you are performing are useful. So show them that for example new KPIs with respect to baseline are better.

$\endgroup$
1
  • $\begingroup$ Yes. And your model needs integration with the real data sources, middleware and/or application UI. And whatever you do: don't just throw it over the wall to sysops. That's when projects fail. Talk to them, regularly, as your model goes into production. $\endgroup$ Commented Feb 9 at 15:44

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.