AI RESEARCH
Robust Personalized Recommendation under Hidden Confounding in MNAR
arXiv CS.LG
•
ArXi:2605.21066v1 Announce Type: new Recommender systems often rely on observational user--item interaction data, which is prone to selection bias due to users' selective interactions with items. Inverse propensity weighting and doubly robust estimators effectively mitigate selection bias under observed confounding, but are unreliable in the presence of hidden confounders.