Professional Machine Learning Engineer

By Google · Question Mode

QuestionQ36

Scaling prototypes into ML models

You work for a global footwear retailer and need to predict when an item will run out of stock using historical inventory data. Customer behavior is highly dynamic because footwear demand is affected by many different factors. You want to serve models trained on all available data, while tracking performance on particular data subsets before promoting them to production. What is the most streamlined and reliable way to carry out this validation?

Explanation

TFX validation uses configured performance metrics and thresholds to assess whether a candidate model is suitable for production, and TensorFlow Model Analysis computes and reports those metrics across specified data slices. This supports production gating while revealing whether performance is acceptable for important subsets rather than only in aggregate. In current TFX, the Evaluator component provides this validation capability; the older ModelValidator component has been folded into it.

Learn more

Community Discussion

No comments yet. Be the first to start the discussion!