Introduces PUID to estimate personalized sensitivity bounds for robust recommendations under hidden confounding in MNAR settings, outperforming global methods on three datasets.
CBPL: A unified calibration and balanc- ing propensity learning framework in causal recommendation for debiasing,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Robust Personalized Recommendation under Hidden Confounding in MNAR
Introduces PUID to estimate personalized sensitivity bounds for robust recommendations under hidden confounding in MNAR settings, outperforming global methods on three datasets.