Spatially varying auto-regressive models for prediction of new human immunodeficiency virus diagnoses

Research output: Contribution to journalArticle

Abstract

In demand of predicting new HIV diagnosis rates based on publicly available HIV data that is abundant in space but has few points in time, we propose a class of spatially varying autoregressive (SVAR) models compounded with conditional autoregressive (CAR) spatial correlation structures. We then propose to use the copula approach and a flexible CAR formulation to model the dependence between adjacent counties. These models allow for spatial and temporal correlation as well as space-time interactions and are naturally suitable for predicting HIV cases and other spatio-temporal disease data that feature a similar data structure. We apply the proposed models to HIV data over Florida, California and New England states and compare them to a range of linear mixed models that have been recently popular for modeling spatio-temporal disease data. The results show that for such data our proposed models outperform the others in terms of prediction.

Original languageEnglish (US)
Pages (from-to)1003-1022
Number of pages20
JournalJournal of the Royal Statistical Society. Series C: Applied Statistics
Volume67
Issue number4
DOIs
StatePublished - Aug 2018

Fingerprint

Autoregressive Model
Virus
Prediction
Spatial Correlation
Spatio-temporal Modeling
Linear Mixed Model
Temporal Correlation
Correlation Structure
Spatial Structure
Copula
Model
Data Structures
Adjacent
Space-time
Human
Autoregressive model
Formulation
Interaction
Range of data

Keywords

  • Bayesian hierarchical models
  • Conditional auto-regressive models
  • Copula
  • Spatiotemporal data

ASJC Scopus subject areas

  • Statistics and Probability
  • Statistics, Probability and Uncertainty

Cite this

@article{c9f4bf4516784d5baa12d85c0a4b0151,
title = "Spatially varying auto-regressive models for prediction of new human immunodeficiency virus diagnoses",
abstract = "In demand of predicting new HIV diagnosis rates based on publicly available HIV data that is abundant in space but has few points in time, we propose a class of spatially varying autoregressive (SVAR) models compounded with conditional autoregressive (CAR) spatial correlation structures. We then propose to use the copula approach and a flexible CAR formulation to model the dependence between adjacent counties. These models allow for spatial and temporal correlation as well as space-time interactions and are naturally suitable for predicting HIV cases and other spatio-temporal disease data that feature a similar data structure. We apply the proposed models to HIV data over Florida, California and New England states and compare them to a range of linear mixed models that have been recently popular for modeling spatio-temporal disease data. The results show that for such data our proposed models outperform the others in terms of prediction.",
keywords = "Bayesian hierarchical models, Conditional auto-regressive models, Copula, Spatiotemporal data",
author = "Lyndsay Shand and Bo Li and Trevor Park and Dolores Albarracin",
year = "2018",
month = "8",
doi = "10.1111/rssc.12269",
language = "English (US)",
volume = "67",
pages = "1003--1022",
journal = "Journal of the Royal Statistical Society. Series C: Applied Statistics",
issn = "0035-9254",
publisher = "Wiley-Blackwell",
number = "4",

}

TY - JOUR

T1 - Spatially varying auto-regressive models for prediction of new human immunodeficiency virus diagnoses

AU - Shand, Lyndsay

AU - Li, Bo

AU - Park, Trevor

AU - Albarracin, Dolores

PY - 2018/8

Y1 - 2018/8

N2 - In demand of predicting new HIV diagnosis rates based on publicly available HIV data that is abundant in space but has few points in time, we propose a class of spatially varying autoregressive (SVAR) models compounded with conditional autoregressive (CAR) spatial correlation structures. We then propose to use the copula approach and a flexible CAR formulation to model the dependence between adjacent counties. These models allow for spatial and temporal correlation as well as space-time interactions and are naturally suitable for predicting HIV cases and other spatio-temporal disease data that feature a similar data structure. We apply the proposed models to HIV data over Florida, California and New England states and compare them to a range of linear mixed models that have been recently popular for modeling spatio-temporal disease data. The results show that for such data our proposed models outperform the others in terms of prediction.

AB - In demand of predicting new HIV diagnosis rates based on publicly available HIV data that is abundant in space but has few points in time, we propose a class of spatially varying autoregressive (SVAR) models compounded with conditional autoregressive (CAR) spatial correlation structures. We then propose to use the copula approach and a flexible CAR formulation to model the dependence between adjacent counties. These models allow for spatial and temporal correlation as well as space-time interactions and are naturally suitable for predicting HIV cases and other spatio-temporal disease data that feature a similar data structure. We apply the proposed models to HIV data over Florida, California and New England states and compare them to a range of linear mixed models that have been recently popular for modeling spatio-temporal disease data. The results show that for such data our proposed models outperform the others in terms of prediction.

KW - Bayesian hierarchical models

KW - Conditional auto-regressive models

KW - Copula

KW - Spatiotemporal data

UR - http://www.scopus.com/inward/record.url?scp=85043579509&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85043579509&partnerID=8YFLogxK

U2 - 10.1111/rssc.12269

DO - 10.1111/rssc.12269

M3 - Article

C2 - 30853848

AN - SCOPUS:85043579509

VL - 67

SP - 1003

EP - 1022

JO - Journal of the Royal Statistical Society. Series C: Applied Statistics

JF - Journal of the Royal Statistical Society. Series C: Applied Statistics

SN - 0035-9254

IS - 4

ER -