This is a bad approach because whenever a new app is added, you might have to retrain your model from scratch, or at least retrain all the components of your model whose number of parameters depends on N (number of apps).
Better method:
In this new framing, whenever there’s a new app you want to consider recommending to a user, you simply need to use new inputs with this new app’s feature instead of having to retrain your model or part of your model from scratch.
Decoupling objectives
What if you have 2 objectives: quality and engagement?
Approach 1
loss = ɑ quality_loss + β engagement_loss
A problem with this approach is that each time you tune α and β—for example, if the quality of your users’ newsfeeds goes up but users’ engagement goes down, you might want to decrease α and increase β—you’ll have to retrain your model.
Approach 2
2 models optimizing for each loss
You can combine the models’ outputs and rank posts by their combined scores: ɑ quality_score + β engagement_score
In general, when there are multiple objectives, it’s a good idea to decouple them first because it makes model development and maintenance easier.
