KAUST DepartmentStatistics Program
Computer, Electrical and Mathematical Sciences and Engineering (CEMSE) Division
Online Publication Date2019-12-26
Print Publication Date2020-06
Embargo End Date2020-12-26
Permanent link to this recordhttp://hdl.handle.net/10754/660876
MetadataShow full item record
AbstractWe set up a general framework for modeling non-Gaussian multivariate stochastic processes by transforming underlying multivariate Gaussian processes. This general framework includes multivariate spatial random felds, multivariate time series, and multivariate spatio-temporal processes, whereas the respective univariate processes can also be seen as special cases. We advocate joint modeling of the transformation and the cross-/auto-correlation structure of the latent multivariate Gaussian process, for better estimation and prediction performance. We provide two useful models, the Tukey g-and-h transformed vector autoregressive model and the sinh-arcsinhtransformed multivariate Matérn random feld. We evaluate them with a simulation study. Finally, we apply the two models to a wind data set for modeling the two perpendicular components of wind speed vectors. Both the simulation study and data analysis show the advantages of the joint modeling approach.
CitationYan, Y., Jeong, J., & Genton, M. G. (2019). Multivariate transformed Gaussian processes. Japanese Journal of Statistics and Data Science. doi:10.1007/s42081-019-00068-6
SponsorsThis publication is based upon work supported by the King Abdullah University of Science and Technology (KAUST) Office of Sponsored Research (OSR) under Award No: OSR-2018-CRG7-3742.
PublisherSpringer Science and Business Media LLC