Andrew B. Lawson 2019 BMTRY 763
|
|
- Dana Underwood
- 5 years ago
- Views:
Transcription
1 BMTRY 763
2 FMD Foot and mouth disease (FMD) is a viral disease of clovenfooted animals and being extremely contagious it spreads rapidly. It reduces animal production, and can sometimes be fatal in young stock. Counts of new cases of FMD were available for parishes within the county of Cumbria, Northern England for the period February 2001-August The data were available as half monthly counts and so there were 13 time periods available for analysis.
3 FMD incidence NW England (Cumbria) 2001 Infected premises (IPs) are to be modeled Within parishes we have counts of IPs and we also have a record of the total number of premises which changes over time. 138 parishes and 13 half-monthly time periods (February 2001-August 2001)
4 FMD first 6 time periods case/pop ratios (row-wise)
5 Data example: FMD in Cumbria Foot and Mouth outbreak in 2001 in Cumbria, UK 138 parishes in area Time period: single biweekly count of Ips Files: FMD_case_parish_data.txt FMD_INLA_Rcode.txt FMD_spatial_worked_examples_Rcode_ABDM.txt
6 FMD (FMDonePERIOD_data.txt) FMDframe<-list(n=138, count=c(1,0,1,0,2,3,1,0,1,0,3,2,0,3,1,6,1,10,2,1,8,4,4,1,1, 3,9,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0, 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,1,0,0,0, 0,0,0,1,1,1,0,0,0,0,0,0,2,0,0,2,0,0,0,0,0,0,0,0,0,0,0,0,0,0, 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,4,0,0)) FMDmap<-readSplus("FMD_splusMAP.txt") plot(fmdmap)
7 Polygon Plot
8 FMD data Counts in parishes (y) Population of premises in parishes (n) Crude rate (y/n) crude rate map [0,0.17)(131) [0.17,0.33)(1) [0.33,0.5)(3) [0.5,0.67)(2) [0.67,0.83)(0) [0.83,1](1)
9 Possible descriptive Model A binomial model may be assumed A logit link also y logit( p ) α v bin( n, p ) i i i i N τ i = α + 1 v 0 (0, ) N(0, τ ) v i
10 INLA code ind<-seq(1:138) formula1<-count~1+f(ind,model="iid,param=c(2,1)) res1<inla(formula1,family="binomial",data=fmdframe,ntrial s=pop,control.compute=list(dic=true)) summary(res1)
11 Results: UH component UH<-res1$summary.random$ind[,2] fillmap(fmdmap,"posterior mean UHcomponent",UH,n.col=4)
12 UH component posterior mean UH component [-0.94,0.64)(108) [0.64,2.23)(9) [2.23,3.81)(13) [3.81,5.4](7)
13 Results: residuals fit<-res1$summary.fitted.values$mean resid<-count-fit fillmap(fmdmap,"estimated crude residuals",resid,n.col=4)
14 Residuals estimated crude residuals [-0.01,2.45)(118) [2.45,4.92)(7) [4.92,7.38)(1) [7.38,9.84](3)
15 Results: infection probability linpred<-res1$summary.linear.predictor$mean prob<-exp(linpred)/(1+exp(linpred)) fillmap(fmdmap,"posterior mean infection probability",prob,n.col=4)
16 posterior mean infection probability Infection probability [0,0.12)(129) [0.12,0.24)(4) [0.24,0.36)(2) [0.36,0.48](3)
17 Results: DIC LOCdic<- res1$dic$local.dic fillmap(fmdmap,"local DIC",LOCdic,n.col=5)
18 DIC local DIC [0.21,1.35)(107) [1.35,2.5)(1) [2.5,3.65)(13) [3.65,4.8)(10) [4.8,5.94](6)
19 Space Time Modeling
20 Space and time considerations Often we need to consider both spatial and temporal effects in disease data Could have a location and date of diagnosis OR could have counts of disease within small areas and fixed time periods The second of these is more common and more aggregated I will only consider this latter situation here
21 Ohio county level respiratory cancer 10 years A well known dataset (full dataset 21 years ) shown here SIRs displayed
22
23 Space-time (ST) Modeling Some notation Assume counts within fixed spatial and temporal periods: map evolutions Both space and time are subscripts in the analysis Consider separable models (with spatial and separate temporal terms) Also interaction effects
24 Notation outcome : y ; RRisk: ij θ ij expected count: e ij i = 1,..., m: small areas j = 1,..., J : time periods
25 Expected Counts Computation (simplest - overall average): e p. y / p = ij ij ij ij i j i j where p ij is the population of the ij th unit
26 Basic retrospective model Infinite population; small disease probability Poisson assumption y ~ Pois( e θ ) ij ij ij log( θ ) = α + S + T + ST S T i j ST ij 0 i j ij : spatial terms : temporal terms ij : interaction
27 Full data set: 21 years of Ohio lung cancer 10 years of SMRs standardized with the statewide rate: Frequently analyzed Row wise from 1979
28
29 Some Random Effect models model 1a: log( θ ) = α + v + u + βt model 1b: ij 0 i i j log( θ ) = α + v + u + γ model 2: ij 0 i i j log( θ ) = α + v + u + γ + γ model 3: ij 0 i i 1j 2 j log( θ ) = α + v + u + γ + ψ model 4: ij 0 i i 2 j ij log( θ ) = α + v + u + γ + γ + ψ ij 0 i i 1j 2 j ij model 5: variants of (3) with ψ ij
30 Simple separable models Spatial components (BYM) plus linear time trend model 1a: log( θ ) = α + v + u + βt t v j i ij 0 i i j : time of j th period τ 1 (0, v ) u N u τ n i N 1... ( δ, u / δ ) i i
31 Simple trend models Spatial components+ random time effect model 1b: log( ) θ α γ = + v + u + ij 0 i i j
32 Random Walk Prior distribution Model 1 b: we assume a random effect for the time element and this has a random walk prior distribution: j 1 1 γ γ N ( γ, τ ) j More generally an AR1 prior could be used: 1 < j 1 γ γ N ( λγ, τ ); 0 λ 1 j
33 Simple space-time models Two random time effects: model 2: log( θ ) = α + v + u + γ + γ γ 1 j ij 0 i i 1j 2 j N 1 (0, τ ) γ j N( 2 j 1, ) γ γ τ γ 2
34 Interaction models Separable models can be enhanced by using additive interaction terms Model 3: log( ) θ α γ ψ = + v + u + + ij 0 i i 2 j ij
35 Interaction priors A variety of priors for the interaction can be assumed (both correlated and non-separable) Knorr-Held (2000) first suggested dependent priors (see Lawson (2013) ch12) Two simple separable examples of possible priors are: ψ ij N(0, τ ) uncorrelated (model 3) ψ ~ N( ψ, τ ) random walk (model 5) ij ψ i, j 1 ψ
36 Interaction models Model 4, as for (3) but with added uncorrelated tiem effect Model 5, as for (3) but with a correlated interaction prior in time model 4: log( θ ) = α + v + u + γ + γ + ψ ij 0 i i 1j 2 j ij model 5: variants of (3) with ψ ij
37 Model fitting Results (WinBUGS) Model DIC pd 1a b
38 Interpretation The temporal trend model does not provide a better fit than the random walk (1a, 1b) The extra RE in model 2 is not needed The inclusion of the interaction in model 3 is significant but model 4 is not good Model 5 with the random walk interaction seems best as it has lowest DIC and smaller pd than model 3
39 Model fitting on INLA Data setup: often space time data is in the form of a matrix: y(i,j), e(i,j) Rows are small areas (eg counties, parishes etc) Columns are time units (eg years, or months days etc) Ohio data: i=1,,88; j=1,,10 Matrix of 88 x 10 dimension For INLA use, it is convenient to reformat the data so that an individual small area represents a row, but the repeated measurements on the small area are repeated rows Long form format
40 R Code For matrices read in with 88 x 10 structures (in R) then the code for conversion is yl<-rep(0,880) el<-rep(0,880) T=10 for (i in 1:88){ for (j in 1:10){ k<-j+t*(i-1) yl[k]<-y[i,j] el[k]<-e[i,j]}}
41 Index setup Indices can now be set up to address different effects: year<-rep(1:10,len=880) region<-rep(1:88,each=10) region2<-region ind2<-rep(1:880)
42 OHIO_Stmapping_INLA_Rcode.txt All models are described in this file. Data is stored in OHIO_STmapping_INLA_RcodeFAL.txt Paste contents of file into R This will yield y, e, and indicator variables in the data.frame called data
43 Models fitted on INLA Spatial only (UH) Spatial only (UH+CH) UH +CH +time trend model 1a UH+CH +time iid UH+CH+time rw1 model 1b UH+CH+time (iid, rw1) model 2 UH +time rw1 +Stint UH+CH+time rw1 +Stint model 3
44 DIC comparison INLA model Model DIC pd WB model 1 Spatial only (UH) UH+CH UH+CH+time trend a 4 UH+CH+time iid UH+CH+time RW1 6 UH+CH+time (iid, rw1) 7 UH+time rw1+st int 8 UH+CH+time rw1+stint b
45 Results: INLA model 7 UH [-0.6,-0.36)(2) [-0.36,-0.12)(16) [-0.12,0.12)(46) [0.12,0.36](23)
46 Results: INLA model 7 time effect yearr time
47 Break
48 FMD modeling in space-time Load data: FMD_case_parish_data.txt INLA code in: FMD_INLA_ST_Rcode.txt
49 Model Development Data is counts within parishes in 13 time periods We also have current population of premises A binomial model may be appropriate y Bin( p, n ) ij ij ij logit( p ) = f ( y ) + f ( predictors) + f (R Es) ij 1 i, j A Poisson model could also be considered if we modulate the Poisson mean with population y ij Pois( μ ) ij log( μ ) = f ( n ) + f ( y ) + f ( predictors and REs) ij 1 ij 2 i, j 1 3
50 Simpler Poisson Model We will look at the Poisson model Also only use random effects Lagged dependencies can be accommodated using the copying facility in INLA. We will only have lagged REs
51 Models Purely spatial (UH+CH) pl offset Purely spatial (UH+CH) pl predictor UH+CH+ pl predictor+ time RW1 UH+CH+ pl offset+ time RW1 UH+CH+ pl predictor+ time RW1+STint
52 Model Description DIC pd Results 1a UH+CH+ pl offset DICs 1b UH+CH+ pl predictor UH+CH+ pl predictor + time RW UH+CH+ pl offset + time RW UH+CH+ pl predictor + time RW1+ STint
53 Model 4 This model has spatial and temporal effects with a pl predictor and ST interaction y ij Pois( μ ) ij μ = exp{ α + α n + v + u + γ + ψ } ij 0 1 ij i i j ij NB lagged effects in count and population not included BUT a lagged dependency random effect is
54 Results UH [-0.73,0.05)(76) [0.05,0.82)(47) [0.82,1.6)(13) [1.6,2.38](1)
55 Results CH [-1.45,-0.62)(31) [-0.62,0.2)(56) [0.2,1.03)(35) [1.03,1.85](16)
56 Results time<-seq(1:13) plot(time,yearr) lines(time,yearr) yearr time
57 Results ST interaction period 1 ST interaction period 1 [-1.07,-0.45)(15) [-0.45,0.17)(93) [0.17,0.8)(10) [0.8,1.42)(6) [1.42,2.05)(4) [2.05,2.67](9)
58 Results: St interaction period 2 ST interaction period 2 [-1.13,-0.44)(13) [-0.44,0.26)(89) [0.26,0.95)(13) [0.95,1.64)(11) [1.64,2.34)(6) [2.34,3.03](6)
59 Finally Other INLA features Measurement error in predictors ( mec, meb) Missingness in outcomes (copy facility) Geographically weighted regression e.g. f(ind,x1,model= besag,graph= ) Smoothed predictors e.g. f(x1,model= rw1 ) Modeling point processes via SPDE facilities (LGCP)
60 INLA OpenBUGS Finally INLA versus OpenBUGS Runs on R x Only through Brugs or R2WinBUGS Large datasets Mixtures x x Posterior functionals? x Special spatial Models Missingness X some: LGCP for point processes Only outcomes in general, but can handle drop-out models X GeoBUGS +CAR models Can handle a range of missingness
61 Cautionary references A study has found that the default prior distributions for precisions in INLA can lead to very inaccurate precision estimates for random effects Carroll, R., Lawson, A. B., Faes, C., Kirby, R. S. and Aregay, M. (2015) Comparing INLA and OpenBUGS for hierarchical Poisson modeling in disease mapping. Spatial and Spatio-temporal Epidemiology, 14-15, See also Taylor, B. and Diggle, P. (2014) INLA or MCMC? A tutorial and comparative evaluation for spatial prediction in log-gaussian Cox Processes. Journal of Statistical Computation and Simulation, 84, 10, Teng, M., Nathoo, F.S., Johnson, T.D. (2017). Bayesian Computation for Log Gaussian Cox Processes: A Comparative Analysis of Methods. Journal of Statistical Computation and Simulation DOI: / So be warned!
62 Conclusions Thanks for your attention! Contact address: INLA examples given in Appendix D of Lawson, A. B. (2013) Bayesian Disease Mapping: hierarchical modeling in spatial epidemiology. 2 nd Ed CRC Press, New York And also chapter 15 of Lawson (2018) 3 rd Ed Full 2 x 1 x 2 day courses on BDM (including WinBUGS, INLA, CARBayes and Nimble) given in MUSC (March) Contacts: MUSC courses June Watson watsonju@musc.edu
63 Andrew B Lawson 2019
How its computed. y outcome data λ parameters hyperparameters. where P denotes the Laplace approximation. k i k k. Andrew B Lawson 2013
Andrew Lawson MUSC INLA INLA s a relatvely new tool that can be used to approxmate posteror dstrbutons n Bayesan models INLA stands for ntegrated Nested Laplace Approxmaton The approxmaton has been known
More informationAnnouncing a Course Sequence: March 9 th - 10 th, March 11 th, and March 12 th - 13 th 2015 Historic Charleston, South Carolina
Announcing a Course Sequence: Introduction to Bayesian Disease Mapping (IBDM) BDM with INLA (BDMI) Advanced Bayesian Disease Mapping (ABDM) March 9 th - 10 th, March 11 th, and March 12 th - 13 th 2015
More informationBayesian Hierarchical Models
Bayesian Hierarchical Models Gavin Shaddick, Millie Green, Matthew Thomas University of Bath 6 th - 9 th December 2016 1/ 34 APPLICATIONS OF BAYESIAN HIERARCHICAL MODELS 2/ 34 OUTLINE Spatial epidemiology
More informationStatistical Analysis of Spatio-temporal Point Process Data. Peter J Diggle
Statistical Analysis of Spatio-temporal Point Process Data Peter J Diggle Department of Medicine, Lancaster University and Department of Biostatistics, Johns Hopkins University School of Public Health
More informationCommunity Health Needs Assessment through Spatial Regression Modeling
Community Health Needs Assessment through Spatial Regression Modeling Glen D. Johnson, PhD CUNY School of Public Health glen.johnson@lehman.cuny.edu Objectives: Assess community needs with respect to particular
More informationBayesian Spatial Health Surveillance
Bayesian Spatial Health Surveillance Allan Clark and Andrew Lawson University of South Carolina 1 Two important problems Clustering of disease: PART 1 Development of Space-time models Modelling vs Testing
More informationA short introduction to INLA and R-INLA
A short introduction to INLA and R-INLA Integrated Nested Laplace Approximation Thomas Opitz, BioSP, INRA Avignon Workshop: Theory and practice of INLA and SPDE November 7, 2018 2/21 Plan for this talk
More informationLocal Likelihood Bayesian Cluster Modeling for small area health data. Andrew Lawson Arnold School of Public Health University of South Carolina
Local Likelihood Bayesian Cluster Modeling for small area health data Andrew Lawson Arnold School of Public Health University of South Carolina Local Likelihood Bayesian Cluster Modelling for Small Area
More informationBayesian dynamic modeling for large space-time weather datasets using Gaussian predictive processes
Bayesian dynamic modeling for large space-time weather datasets using Gaussian predictive processes Andrew O. Finley 1 and Sudipto Banerjee 2 1 Department of Forestry & Department of Geography, Michigan
More informationBayesian dynamic modeling for large space-time weather datasets using Gaussian predictive processes
Bayesian dynamic modeling for large space-time weather datasets using Gaussian predictive processes Alan Gelfand 1 and Andrew O. Finley 2 1 Department of Statistical Science, Duke University, Durham, North
More informationBayesian dynamic modeling for large space-time weather datasets using Gaussian predictive processes
Bayesian dynamic modeling for large space-time weather datasets using Gaussian predictive processes Sudipto Banerjee 1 and Andrew O. Finley 2 1 Biostatistics, School of Public Health, University of Minnesota,
More informationPoint process with spatio-temporal heterogeneity
Point process with spatio-temporal heterogeneity Jony Arrais Pinto Jr Universidade Federal Fluminense Universidade Federal do Rio de Janeiro PASI June 24, 2014 * - Joint work with Dani Gamerman and Marina
More informationBayesian dynamic modeling for large space-time weather datasets using Gaussian predictive processes
Bayesian dynamic modeling for large space-time weather datasets using Gaussian predictive processes Andrew O. Finley Department of Forestry & Department of Geography, Michigan State University, Lansing
More informationModelling spatio-temporal patterns of disease
Modelling spatio-temporal patterns of disease Peter J Diggle CHICAS combining health information, computation and statistics References AEGISS Brix, A. and Diggle, P.J. (2001). Spatio-temporal prediction
More informationRonald Christensen. University of New Mexico. Albuquerque, New Mexico. Wesley Johnson. University of California, Irvine. Irvine, California
Texts in Statistical Science Bayesian Ideas and Data Analysis An Introduction for Scientists and Statisticians Ronald Christensen University of New Mexico Albuquerque, New Mexico Wesley Johnson University
More informationBayesian Areal Wombling for Geographic Boundary Analysis
Bayesian Areal Wombling for Geographic Boundary Analysis Haolan Lu, Haijun Ma, and Bradley P. Carlin haolanl@biostat.umn.edu, haijunma@biostat.umn.edu, and brad@biostat.umn.edu Division of Biostatistics
More informationCluster investigations using Disease mapping methods International workshop on Risk Factors for Childhood Leukemia Berlin May
Cluster investigations using Disease mapping methods International workshop on Risk Factors for Childhood Leukemia Berlin May 5-7 2008 Peter Schlattmann Institut für Biometrie und Klinische Epidemiologie
More informationHierarchical modelling of performance indicators, with application to MRSA & teenage conception rates
Hierarchical modelling of performance indicators, with application to MRSA & teenage conception rates Hayley E Jones School of Social and Community Medicine, University of Bristol, UK Thanks to David Spiegelhalter,
More informationR-INLA. Sam Clifford Su-Yun Kang Jeff Hsieh. 30 August Bayesian Research and Analysis Group 1 / 14
1 / 14 R-INLA Sam Clifford Su-Yun Kang Jeff Hsieh Bayesian Research and Analysis Group 30 August 2012 What is R-INLA? R package for Bayesian computation Integrated Nested Laplace Approximation MCMC free
More informationSTA 216, GLM, Lecture 16. October 29, 2007
STA 216, GLM, Lecture 16 October 29, 2007 Efficient Posterior Computation in Factor Models Underlying Normal Models Generalized Latent Trait Models Formulation Genetic Epidemiology Illustration Structural
More informationGaussian Process Regression Model in Spatial Logistic Regression
Journal of Physics: Conference Series PAPER OPEN ACCESS Gaussian Process Regression Model in Spatial Logistic Regression To cite this article: A Sofro and A Oktaviarina 018 J. Phys.: Conf. Ser. 947 01005
More informationDisease mapping with Gaussian processes
EUROHEIS2 Kuopio, Finland 17-18 August 2010 Aki Vehtari (former Helsinki University of Technology) Department of Biomedical Engineering and Computational Science (BECS) Acknowledgments Researchers - Jarno
More informationGeneralized common spatial factor model
Biostatistics (2003), 4, 4,pp. 569 582 Printed in Great Britain Generalized common spatial factor model FUJUN WANG Eli Lilly and Company, Indianapolis, IN 46285, USA MELANIE M. WALL Division of Biostatistics,
More informationLecture 23. Spatio-temporal Models. Colin Rundel 04/17/2017
Lecture 23 Spatio-temporal Models Colin Rundel 04/17/2017 1 Spatial Models with AR time dependence 2 Example - Weather station data Based on Andrew Finley and Sudipto Banerjee s notes from National Ecological
More informationWEB application for the analysis of spatio-temporal data
EXTERNAL SCIENTIFIC REPORT APPROVED: 17 October 2016 WEB application for the analysis of spatio-temporal data Abstract Machteld Varewyck (Open Analytics NV), Tobias Verbeke (Open Analytics NV) In specific
More informationPumps, Maps and Pea Soup: Spatio-temporal methods in environmental epidemiology
Pumps, Maps and Pea Soup: Spatio-temporal methods in environmental epidemiology Gavin Shaddick Department of Mathematical Sciences University of Bath 2012-13 van Eeden lecture Thanks Constance van Eeden
More informationApproaches for Multiple Disease Mapping: MCAR and SANOVA
Approaches for Multiple Disease Mapping: MCAR and SANOVA Dipankar Bandyopadhyay Division of Biostatistics, University of Minnesota SPH April 22, 2015 1 Adapted from Sudipto Banerjee s notes SANOVA vs MCAR
More informationModels for spatial data (cont d) Types of spatial data. Types of spatial data (cont d) Hierarchical models for spatial data
Hierarchical models for spatial data Based on the book by Banerjee, Carlin and Gelfand Hierarchical Modeling and Analysis for Spatial Data, 2004. We focus on Chapters 1, 2 and 5. Geo-referenced data arise
More informationFrailty Modeling for Spatially Correlated Survival Data, with Application to Infant Mortality in Minnesota By: Sudipto Banerjee, Mela. P.
Frailty Modeling for Spatially Correlated Survival Data, with Application to Infant Mortality in Minnesota By: Sudipto Banerjee, Melanie M. Wall, Bradley P. Carlin November 24, 2014 Outlines of the talk
More informationMultivariate spatial modeling
Multivariate spatial modeling Point-referenced spatial data often come as multivariate measurements at each location Chapter 7: Multivariate Spatial Modeling p. 1/21 Multivariate spatial modeling Point-referenced
More informationBayesian Inference on Joint Mixture Models for Survival-Longitudinal Data with Multiple Features. Yangxin Huang
Bayesian Inference on Joint Mixture Models for Survival-Longitudinal Data with Multiple Features Yangxin Huang Department of Epidemiology and Biostatistics, COPH, USF, Tampa, FL yhuang@health.usf.edu January
More informationRepresent processes and observations that span multiple levels (aka multi level models) R 2
Hierarchical models Hierarchical models Represent processes and observations that span multiple levels (aka multi level models) R 1 R 2 R 3 N 1 N 2 N 3 N 4 N 5 N 6 N 7 N 8 N 9 N i = true abundance on a
More informationAggregated cancer incidence data: spatial models
Aggregated cancer incidence data: spatial models 5 ième Forum du Cancéropôle Grand-est - November 2, 2011 Erik A. Sauleau Department of Biostatistics - Faculty of Medicine University of Strasbourg ea.sauleau@unistra.fr
More informationBayesian SAE using Complex Survey Data Lecture 4A: Hierarchical Spatial Bayes Modeling
Bayesian SAE using Complex Survey Data Lecture 4A: Hierarchical Spatial Bayes Modeling Jon Wakefield Departments of Statistics and Biostatistics University of Washington 1 / 37 Lecture Content Motivation
More informationSpace-time modelling of air pollution with array methods
Space-time modelling of air pollution with array methods Dae-Jin Lee Royal Statistical Society Conference Edinburgh 2009 D.-J. Lee (Uc3m) GLAM: Array methods in Statistics RSS 09 - Edinburgh # 1 Motivation
More informationPerformance of INLA analysing bivariate meta-regression and age-period-cohort models
Performance of INLA analysing bivariate meta-regression and age-period-cohort models Andrea Riebler Biostatistics Unit, Institute of Social and Preventive Medicine University of Zurich INLA workshop, May
More informationTo link to this article: PLEASE SCROLL DOWN FOR ARTICLE
This article was downloaded by: [University of West Florida] On: 25 November 2014, At: 10:11 Publisher: Taylor & Francis Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered
More informationSpatio-Temporal Threshold Models for Relating UV Exposures and Skin Cancer in the Central United States
Spatio-Temporal Threshold Models for Relating UV Exposures and Skin Cancer in the Central United States Laura A. Hatfield and Bradley P. Carlin Division of Biostatistics School of Public Health University
More informationAnalysis of Marked Point Patterns with Spatial and Non-spatial Covariate Information
Analysis of Marked Point Patterns with Spatial and Non-spatial Covariate Information p. 1/27 Analysis of Marked Point Patterns with Spatial and Non-spatial Covariate Information Shengde Liang, Bradley
More informationMultivariate Count Time Series Modeling of Surveillance Data
Multivariate Count Time Series Modeling of Surveillance Data Leonhard Held 1 Michael Höhle 2 1 Epidemiology, Biostatistics and Prevention Institute, University of Zurich, Switzerland 2 Department of Mathematics,
More informationLog Gaussian Cox Processes. Chi Group Meeting February 23, 2016
Log Gaussian Cox Processes Chi Group Meeting February 23, 2016 Outline Typical motivating application Introduction to LGCP model Brief overview of inference Applications in my work just getting started
More informationARIC Manuscript Proposal # PC Reviewed: _9/_25_/06 Status: A Priority: _2 SC Reviewed: _9/_25_/06 Status: A Priority: _2
ARIC Manuscript Proposal # 1186 PC Reviewed: _9/_25_/06 Status: A Priority: _2 SC Reviewed: _9/_25_/06 Status: A Priority: _2 1.a. Full Title: Comparing Methods of Incorporating Spatial Correlation in
More informationHierarchical Modeling for Spatio-temporal Data
Hierarchical Modeling for Spatio-temporal Data Sudipto Banerjee 1 and Andrew O. Finley 2 1 Biostatistics, School of Public Health, University of Minnesota, Minneapolis, Minnesota, U.S.A. 2 Department of
More informationDIC: Deviance Information Criterion
(((( Welcome Page Latest News DIC: Deviance Information Criterion Contact us/bugs list WinBUGS New WinBUGS examples FAQs DIC GeoBUGS DIC (Deviance Information Criterion) is a Bayesian method for model
More informationFlexible Spatio-temporal smoothing with array methods
Int. Statistical Inst.: Proc. 58th World Statistical Congress, 2011, Dublin (Session IPS046) p.849 Flexible Spatio-temporal smoothing with array methods Dae-Jin Lee CSIRO, Mathematics, Informatics and
More informationRobust MCMC Algorithms for Bayesian Inference in Stochastic Epidemic Models.
Robust MCMC Algorithms for Bayesian Inference in Stochastic Epidemic Models. An Application to the 2001 UK Foot-and-Mouth Outbreak Theodore Kypraios @ University of Nottingham Gareth O. Roberts @ Lancaster
More informationFaculty of Health Sciences. Regression models. Counts, Poisson regression, Lene Theil Skovgaard. Dept. of Biostatistics
Faculty of Health Sciences Regression models Counts, Poisson regression, 27-5-2013 Lene Theil Skovgaard Dept. of Biostatistics 1 / 36 Count outcome PKA & LTS, Sect. 7.2 Poisson regression The Binomial
More informationLecture 8. Poisson models for counts
Lecture 8. Poisson models for counts Jesper Rydén Department of Mathematics, Uppsala University jesper.ryden@math.uu.se Statistical Risk Analysis Spring 2014 Absolute risks The failure intensity λ(t) describes
More informationBayesian Statistics Part III: Building Bayes Theorem Part IV: Prior Specification
Bayesian Statistics Part III: Building Bayes Theorem Part IV: Prior Specification Michael Anderson, PhD Hélène Carabin, DVM, PhD Department of Biostatistics and Epidemiology The University of Oklahoma
More informationHierarchical Modelling for Univariate Spatial Data
Hierarchical Modelling for Univariate Spatial Data Sudipto Banerjee 1 and Andrew O. Finley 2 1 Biostatistics, School of Public Health, University of Minnesota, Minneapolis, Minnesota, U.S.A. 2 Department
More informationMeasurement Error in Spatial Modeling of Environmental Exposures
Measurement Error in Spatial Modeling of Environmental Exposures Chris Paciorek, Alexandros Gryparis, and Brent Coull August 9, 2005 Department of Biostatistics Harvard School of Public Health www.biostat.harvard.edu/~paciorek
More informationBayesian Regression Linear and Logistic Regression
When we want more than point estimates Bayesian Regression Linear and Logistic Regression Nicole Beckage Ordinary Least Squares Regression and Lasso Regression return only point estimates But what if we
More informationModels for Count and Binary Data. Poisson and Logistic GWR Models. 24/07/2008 GWR Workshop 1
Models for Count and Binary Data Poisson and Logistic GWR Models 24/07/2008 GWR Workshop 1 Outline I: Modelling counts Poisson regression II: Modelling binary events Logistic Regression III: Poisson Regression
More informationEstimating the long-term health impact of air pollution using spatial ecological studies. Duncan Lee
Estimating the long-term health impact of air pollution using spatial ecological studies Duncan Lee EPSRC and RSS workshop 12th September 2014 Acknowledgements This is joint work with Alastair Rushworth
More informationINLA for Spatial Statistics
INLA for Spatial Statistics 5. Log-Gaussian Co Processes Daniel Simpson Department of Mathematical Sciences University of Bath Outline spatial point processes fitting log Gaussian Co processes eamples
More informationSummary STK 4150/9150
STK4150 - Intro 1 Summary STK 4150/9150 Odd Kolbjørnsen May 22 2017 Scope You are expected to know and be able to use basic concepts introduced in the book. You knowledge is expected to be larger than
More informationHierarchical Nearest-Neighbor Gaussian Process Models for Large Geo-statistical Datasets
Hierarchical Nearest-Neighbor Gaussian Process Models for Large Geo-statistical Datasets Abhirup Datta 1 Sudipto Banerjee 1 Andrew O. Finley 2 Alan E. Gelfand 3 1 University of Minnesota, Minneapolis,
More informationHierarchical Modeling for Univariate Spatial Data
Hierarchical Modeling for Univariate Spatial Data Geography 890, Hierarchical Bayesian Models for Environmental Spatial Data Analysis February 15, 2011 1 Spatial Domain 2 Geography 890 Spatial Domain This
More informationOrder-q stochastic processes. Bayesian nonparametric applications
Order-q dependent stochastic processes in Bayesian nonparametric applications Department of Statistics, ITAM, Mexico BNP 2015, Raleigh, NC, USA 25 June, 2015 Contents Order-1 process Application in survival
More informationDIC, AIC, BIC, PPL, MSPE Residuals Predictive residuals
DIC, AIC, BIC, PPL, MSPE Residuals Predictive residuals Overall Measures of GOF Deviance: this measures the overall likelihood of the model given a parameter vector D( θ) = 2 log L( θ) This can be averaged
More informationSpacetime models in R-INLA. Elias T. Krainski
Spacetime models in R-INLA Elias T. Krainski 2 Outline Separable space-time models Infant mortality in Paraná PM-10 concentration in Piemonte, Italy 3 Multivariate dynamic regression model y t : n observations
More informationSpatio-temporal modeling of weekly malaria incidence in children under 5 for early epidemic detection in Mozambique
Spatio-temporal modeling of weekly malaria incidence in children under 5 for early epidemic detection in Mozambique Katie Colborn, PhD Department of Biostatistics and Informatics University of Colorado
More informationChapter 4 - Fundamentals of spatial processes Lecture notes
Chapter 4 - Fundamentals of spatial processes Lecture notes Geir Storvik January 21, 2013 STK4150 - Intro 2 Spatial processes Typically correlation between nearby sites Mostly positive correlation Negative
More informationCluster Analysis using SaTScan
Cluster Analysis using SaTScan Summary 1. Statistical methods for spatial epidemiology 2. Cluster Detection What is a cluster? Few issues 3. Spatial and spatio-temporal Scan Statistic Methods Probability
More informationFully Bayesian Spatial Analysis of Homicide Rates.
Fully Bayesian Spatial Analysis of Homicide Rates. Silvio A. da Silva, Luiz L.M. Melo and Ricardo S. Ehlers Universidade Federal do Paraná, Brazil Abstract Spatial models have been used in many fields
More informationLISA Short Course Series Generalized Linear Models (GLMs) & Categorical Data Analysis (CDA) in R. Liang (Sally) Shan Nov. 4, 2014
LISA Short Course Series Generalized Linear Models (GLMs) & Categorical Data Analysis (CDA) in R Liang (Sally) Shan Nov. 4, 2014 L Laboratory for Interdisciplinary Statistical Analysis LISA helps VT researchers
More informationBayesian non-parametric model to longitudinally predict churn
Bayesian non-parametric model to longitudinally predict churn Bruno Scarpa Università di Padova Conference of European Statistics Stakeholders Methodologists, Producers and Users of European Statistics
More informationLecture 5: Poisson and logistic regression
Dankmar Böhning Southampton Statistical Sciences Research Institute University of Southampton, UK S 3 RI, 3-5 March 2014 introduction to Poisson regression application to the BELCAP study introduction
More informationChallenges in modelling air pollution and understanding its impact on human health
Challenges in modelling air pollution and understanding its impact on human health Alastair Rushworth Joint Statistical Meeting, Seattle Wednesday August 12 th, 2015 Acknowledgements Work in this talk
More information2015 SISG Bayesian Statistics for Genetics R Notes: Generalized Linear Modeling
2015 SISG Bayesian Statistics for Genetics R Notes: Generalized Linear Modeling Jon Wakefield Departments of Statistics and Biostatistics, University of Washington 2015-07-24 Case control example We analyze
More informationAlain F. Zuur Highland Statistics Ltd. Newburgh, UK.
Annex C: Analysis of 7 wind farm data sets 1.1 Introduction Alain F. Zuur Highland Statistics Ltd. Newburgh, UK. www.highstat.com Required knowledge for this text is data exploration, multiple linear regression,
More informationSpatial modelling using INLA
Spatial modelling using INLA August 30, 2012 Bayesian Research and Analysis Group Queensland University of Technology Log Gaussian Cox process Hierarchical Poisson process λ(s) = exp(z(s)) for Z(s) a Gaussian
More informationDisease mapping with Gaussian processes
Liverpool, UK, 4 5 November 3 Aki Vehtari Department of Biomedical Engineering and Computational Science (BECS) Outline Example: Alcohol related deaths in Finland Spatial priors and benefits of GP prior
More information-A wild house sparrow population case study
Bayesian animal model using Integrated Nested Laplace Approximations -A wild house sparrow population case study Anna M. Holand Ingelin Steinsland Animal model workshop with application to ecology at Oulu
More informationGaussian processes for spatial modelling in environmental health: parameterizing for flexibility vs. computational efficiency
Gaussian processes for spatial modelling in environmental health: parameterizing for flexibility vs. computational efficiency Chris Paciorek March 11, 2005 Department of Biostatistics Harvard School of
More informationBayesian Dynamic Modeling for Space-time Data in R
Bayesian Dynamic Modeling for Space-time Data in R Andrew O. Finley and Sudipto Banerjee September 5, 2014 We make use of several libraries in the following example session, including: ˆ library(fields)
More informationConcepts and Applications of Kriging. Eric Krause
Concepts and Applications of Kriging Eric Krause Sessions of note Tuesday ArcGIS Geostatistical Analyst - An Introduction 8:30-9:45 Room 14 A Concepts and Applications of Kriging 10:15-11:30 Room 15 A
More informationChapter 4 - Spatial processes R packages and software Lecture notes
TK4150 - Intro 1 Chapter 4 - Spatial processes R packages and software Lecture notes Odd Kolbjørnsen and Geir Storvik February 13, 2017 STK4150 - Intro 2 Last time General set up for Kriging type problems
More informationComparison of Bayesian Spatio-Temporal Models for Chronic Diseases
Journal of Data Science 8(2010), 189-211 Comparison of Bayesian Spatio-Temporal Models for Chronic Diseases Hoon Kim 1 and HeeJeong Lim 2 1 California State Polytechnic University and 2 California State
More informationExample using R: Heart Valves Study
Example using R: Heart Valves Study Goal: Show that the thrombogenicity rate (TR) is less than two times the objective performance criterion R and WinBUGS Examples p. 1/27 Example using R: Heart Valves
More informationPart 8: GLMs and Hierarchical LMs and GLMs
Part 8: GLMs and Hierarchical LMs and GLMs 1 Example: Song sparrow reproductive success Arcese et al., (1992) provide data on a sample from a population of 52 female song sparrows studied over the course
More informationLongitudinal Data Analysis Using Stata Paul D. Allison, Ph.D. Upcoming Seminar: May 18-19, 2017, Chicago, Illinois
Longitudinal Data Analysis Using Stata Paul D. Allison, Ph.D. Upcoming Seminar: May 18-19, 217, Chicago, Illinois Outline 1. Opportunities and challenges of panel data. a. Data requirements b. Control
More informationSTAT Lecture 11: Bayesian Regression
STAT 491 - Lecture 11: Bayesian Regression Generalized Linear Models Generalized linear models (GLMs) are a class of techniques that include linear regression, logistic regression, and Poisson regression.
More informationNon-parametric Bayesian Modeling and Fusion of Spatio-temporal Information Sources
th International Conference on Information Fusion Chicago, Illinois, USA, July -8, Non-parametric Bayesian Modeling and Fusion of Spatio-temporal Information Sources Priyadip Ray Department of Electrical
More informationStatistics 910, #15 1. Kalman Filter
Statistics 910, #15 1 Overview 1. Summary of Kalman filter 2. Derivations 3. ARMA likelihoods 4. Recursions for the variance Kalman Filter Summary of Kalman filter Simplifications To make the derivations
More informationModelling geoadditive survival data
Modelling geoadditive survival data Thomas Kneib & Ludwig Fahrmeir Department of Statistics, Ludwig-Maximilians-University Munich 1. Leukemia survival data 2. Structured hazard regression 3. Mixed model
More informationarxiv: v1 [stat.me] 13 May 2017
A note on intrinsic Conditional Autoregressive models for disconnected graphs Anna Freni-Sterrantino Massimo Ventrucci arxiv:1705.04854v1 [stat.me] 13 May 2017 Håvard Rue Abstract In this note we discuss
More informationPart 7: Hierarchical Modeling
Part 7: Hierarchical Modeling!1 Nested data It is common for data to be nested: i.e., observations on subjects are organized by a hierarchy Such data are often called hierarchical or multilevel For example,
More informationSpatio-Temporal Modelling of Credit Default Data
1/20 Spatio-Temporal Modelling of Credit Default Data Sathyanarayan Anand Advisor: Prof. Robert Stine The Wharton School, University of Pennsylvania April 29, 2011 2/20 Outline 1 Background 2 Conditional
More informationDirichlet process Bayesian clustering with the R package PReMiuM
Dirichlet process Bayesian clustering with the R package PReMiuM Dr Silvia Liverani Brunel University London July 2015 Silvia Liverani (Brunel University London) Profile Regression 1 / 18 Outline Motivation
More informationBayesian Nonparametric Spatio-Temporal Models for Disease Incidence Data
Bayesian Nonparametric Spatio-Temporal Models for Disease Incidence Data Athanasios Kottas 1, Jason A. Duan 2, Alan E. Gelfand 3 University of California, Santa Cruz 1 Yale School of Management 2 Duke
More informationBayesian hierarchical space time model applied to highresolution hindcast data of significant wave height
Bayesian hierarchical space time model applied to highresolution hindcast data of Presented at Wave workshop 2013, Banff, Canada Erik Vanem Introduction The Bayesian hierarchical space-time models were
More informationHierarchical Modeling for Multivariate Spatial Data
Hierarchical Modeling for Multivariate Spatial Data Sudipto Banerjee 1 and Andrew O. Finley 2 1 Biostatistics, School of Public Health, University of Minnesota, Minneapolis, Minnesota, U.S.A. 2 Department
More informationscrna-seq Differential expression analysis methods Olga Dethlefsen NBIS, National Bioinformatics Infrastructure Sweden October 2017
scrna-seq Differential expression analysis methods Olga Dethlefsen NBIS, National Bioinformatics Infrastructure Sweden October 2017 Olga (NBIS) scrna-seq de October 2017 1 / 34 Outline Introduction: what
More informationGaussian copula regression using R
University of Warwick August 17, 2011 Gaussian copula regression using R Guido Masarotto University of Padua Cristiano Varin Ca Foscari University, Venice 1/ 21 Framework Gaussian copula marginal regression
More informationA tutorial in spatial and spatio-temporal models with R-INLA
A tutorial in spatial and spatio-temporal models with R-INLA Marta Blangiardo 1, Michela Cameletti 2, Gianluca Baio 3,4, Håvard Rue 5 1 MRC-HPA Centre for Environment and Health, Department of Epidemiology
More informationLecture 14 Bayesian Models for Spatio-Temporal Data
Lecture 14 Bayesian Models for Spatio-Temporal Data Dennis Sun Stats 253 August 13, 2014 Outline of Lecture 1 Recap of Bayesian Models 2 Empirical Bayes 3 Case 1: Long-Lead Forecasting of Sea Surface Temperatures
More informationCOS513 LECTURE 8 STATISTICAL CONCEPTS
COS513 LECTURE 8 STATISTICAL CONCEPTS NIKOLAI SLAVOV AND ANKUR PARIKH 1. MAKING MEANINGFUL STATEMENTS FROM JOINT PROBABILITY DISTRIBUTIONS. A graphical model (GM) represents a family of probability distributions
More informationI don t have much to say here: data are often sampled this way but we more typically model them in continuous space, or on a graph
Spatial analysis Huge topic! Key references Diggle (point patterns); Cressie (everything); Diggle and Ribeiro (geostatistics); Dormann et al (GLMMs for species presence/abundance); Haining; (Pinheiro and
More informationMultivariate Survival Analysis
Multivariate Survival Analysis Previously we have assumed that either (X i, δ i ) or (X i, δ i, Z i ), i = 1,..., n, are i.i.d.. This may not always be the case. Multivariate survival data can arise in
More information