When labeling we have two pre-defined labels:
- Skip (
? in TSV) to skip the product from the prediction (it won't be sampled next time).
- Reject (
! in TSV) to tell the algorithm that and perhaps its category, are not present in the target shop (the algorithm will try to learn other similar products that might be not present in a shop)
It's hard for me to tell by the name what Reject is.
Perhaps a better name would be:
- missing
- absent
- not present
Perhaps also we should also put some explanations on the screen (e.g. when you hover over)