In this work we propose and show a general framework of distribution-free predictive inference for Spatio-Functional Regression models using Conformal Prediction techniques. In particular we focus on Geographically Weighted Regression (GWR) and Heteroskedastic Geographically Weighted Regression (HGWR) with two novel choices of non-conformity measures to compute simultaneous prediction bands.
Conformal prediction for spatial functional regression models
Andrea Diana
Membro del Collaboration Group
;Elvira RomanoMembro del Collaboration Group
;Antonio IrpinoMembro del Collaboration Group
2022
Abstract
In this work we propose and show a general framework of distribution-free predictive inference for Spatio-Functional Regression models using Conformal Prediction techniques. In particular we focus on Geographically Weighted Regression (GWR) and Heteroskedastic Geographically Weighted Regression (HGWR) with two novel choices of non-conformity measures to compute simultaneous prediction bands.File in questo prodotto:
Non ci sono file associati a questo prodotto.
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
