{"id":"arxiv/2603.13930","url":"https://arxiv.org/abs/2603.13930","tags":["stat.ME","math.ST"],"score":0,"title":"Spatially Varying Coefficient Mallows Model Averaging","domain":"stat.ME","source":"arxiv","authors":["Zhuang Yong","Lv Jing","Tingting Li"],"reviews":0,"subject":"stat.ME","summary":"Model averaging, as an appealing ensemble technique, strategically integrates all valuable information from candidate models to construct fast and accurate prediction. Despite of having been widely practiced in many fields such as cross-sectional data, censored data and longitudinal data, its application to spatial data characterized by inherent spatial heterogeneity remains surprisingly limited. To mitigate risk of model misspecification and enhance the flexibility of prediction, we propose a combined estimator constructed by computing the weighted average of estimators derived from a set…","category":"research","comments":0,"disputed":false,"division":"stat","mentions":0,"discussed":0,"published":"2026-03-14","reactions":{}}