Spatial Prediction for Multivariate Non-Gaussian Data

Abstract

With the ever increasing volume of geo-referenced datasets, there is a real need for better statistical estimation and prediction techniques for spatial analysis. Most existing approaches focus on predicting multivariate Gaussian spatial processes, but as the data may consist of non-Gaussian (or mixed type) variables, this creates two challenges: (1) how to accurately capture the dependencies among different data types, both Gaussian and non-Gaussian; and (2) how to efficiently predict multivariate non-Gaussian spatial processes. In this article, we propose a generic approach for predicting multiple response variables of mixed types. The proposed approach accurately captures cross-spatial dependencies among response variables and reduces the computational burden by projecting the spatial process to a lower dimensional space with knot-based techniques. Efficient approximations are provided to estimate posterior marginals of latent variables for the predictive process, and extensive experimental evaluations based on both simulation and real-life datasets are provided to demonstrate the effectiveness and efficiency of this new approach.

Joint spatial modelling of malaria incidence and vector's abundance shows heterogeneity in malaria‐vector geographical relationships

“…One of these dependencies is the explicit spatial correlation, that is malaria and vectors have their own spatial variation, part of which may be considered shared between them. Understanding this shared variation is the focus of this study with the aim to improve targeted surveillance and control by delineating areas that can be prioritised for interventions based on the number of vectors involved and their relative risk, and areas in need of future sampling and investigation in terms of local factors (Finley et al., 2014; Liu et al., 2017) that were missed in the modelling but that contributes at the local malaria transmission. The identification of this shared component is mostly neglected in previous literature, but is key for detailed understanding of the spatial patterns in the interactions between malaria and vectors.…”