Have been learning Machine Learning concepts and doing hands-on past few months. Now got a bit different doubt than implementation of code.
This is regarding measures to be considered while deploying in a Prod environment, say something like "A newbie's checklist to ML production" example -
- Are Feedback loops always considered?
- Is cross-validation done in final prod model or only during train/test the model?
- How the auto-suggestion in Amazon or any ecommerce website works? - Meaning does the model get trained on every page load or already pre-trained and just predicts (suggests) based on user selections?
I know the questions are too general and newbie, but will help many to understand the concepts relating to real-world. Please advise.
Let me know for any further information and I will be happy to furnish/respond.