{"id":"arxiv/2603.17599","url":"https://arxiv.org/abs/2603.17599","tags":["stat.ME","math.ST","stat.AP"],"score":0,"title":"Prediction with Missing Data: Target Probabilities and Missingness Mechanisms","domain":"stat.ME","source":"arxiv","authors":["Pierre Catoire","Robin Genuer","Cecile Proust-Lima"],"reviews":0,"subject":"stat.ME","summary":"Conditions ensuring optimal parameter estimation in the presence of missing data are well established in inference, typically relying on the Missing-at-Random (MAR) assumption. In prediction, similar principles are often assumed to apply. However, methods considered biased in inference, such as pattern sub-modelling or unconditional imputation, have been shown to achieve optimal predictive performance under any missingness mechanism, including non-MAR (MNAR). To explain this apparent contradiction, we introduce a new formal framework for describing missingness in prediction. Central to this…","category":"research","comments":0,"disputed":false,"division":"stat","mentions":0,"discussed":0,"published":"2026-03-18","reactions":{}}