Download pdf - DATA ANALYSIS USING HIERARCHICAL GENERALIZED LINEAR MODELS

Transcript
Page 1: DATA ANALYSIS USING HIERARCHICAL GENERALIZED LINEAR MODELS

DATA ANALYSIS USINGHIERARCHICAL GENERALIZED

LINEAR MODELS WITH R

Youngjo Lee, Lars RΓΆnnegΓ₯rd & Maengseok Noh

Page 2: DATA ANALYSIS USING HIERARCHICAL GENERALIZED LINEAR MODELS

What did John Nelder achieve in the 1970’s?

Page 3: DATA ANALYSIS USING HIERARCHICAL GENERALIZED LINEAR MODELS

Gaussian

Poisson Binomial

Gamma

𝐿𝐿 = �𝑖𝑖=1

𝑛𝑛12πœ‹πœ‹πœŽπœŽ

𝑒𝑒12𝜎𝜎2 π‘¦π‘¦π‘–π‘–βˆ’πœ‡πœ‡ 2

πœ‡πœ‡ = 𝑋𝑋𝑋𝑋

𝐿𝐿 = �𝑖𝑖=1

π‘›π‘›πœ‡πœ‡π‘¦π‘¦π‘–π‘–π‘’π‘’βˆ’πœ‡πœ‡π‘¦π‘¦π‘–π‘–!

log(πœ‡πœ‡) = 𝑋𝑋𝑋𝑋

𝐿𝐿 = �𝑖𝑖=1

𝑛𝑛𝑛𝑛𝑦𝑦𝑖𝑖 πœ‡πœ‡π‘¦π‘¦π‘–π‘–(1 βˆ’ πœ‡πœ‡)π‘›π‘›βˆ’π‘¦π‘¦π‘–π‘–

𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙(πœ‡πœ‡) = 𝑋𝑋𝑋𝑋

𝐿𝐿 = �𝑖𝑖=1

𝑛𝑛1

Ξ“(π‘˜π‘˜) πœƒπœƒπ‘˜π‘˜π‘¦π‘¦π‘–π‘–π‘˜π‘˜βˆ’1𝑒𝑒

βˆ’π‘¦π‘¦π‘–π‘–πœƒπœƒ

kπœƒπœƒ ≑ πœ‡πœ‡; 1/πœ‡πœ‡ = 𝑋𝑋𝑋𝑋

Page 4: DATA ANALYSIS USING HIERARCHICAL GENERALIZED LINEAR MODELS

Gaussian

Poisson Binomial

Gamma

𝐿𝐿 = �𝑖𝑖=1

𝑛𝑛12πœ‹πœ‹πœŽπœŽ

𝑒𝑒12𝜎𝜎2 π‘¦π‘¦π‘–π‘–βˆ’πœ‡πœ‡ 2

πœ‡πœ‡ = 𝑋𝑋𝑋𝑋

𝐿𝐿 = �𝑖𝑖=1

π‘›π‘›πœ‡πœ‡π‘¦π‘¦π‘–π‘–π‘’π‘’βˆ’πœ‡πœ‡π‘¦π‘¦π‘–π‘–!

log(πœ‡πœ‡) = 𝑋𝑋𝑋𝑋

𝐿𝐿 = �𝑖𝑖=1

𝑛𝑛𝑛𝑛𝑦𝑦𝑖𝑖 πœ‡πœ‡π‘¦π‘¦π‘–π‘–(1 βˆ’ πœ‡πœ‡)π‘›π‘›βˆ’π‘¦π‘¦π‘–π‘–

𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙(πœ‡πœ‡) = 𝑋𝑋𝑋𝑋

𝐿𝐿 = �𝑖𝑖=1

𝑛𝑛1

Ξ“(π‘˜π‘˜) πœƒπœƒπ‘˜π‘˜π‘¦π‘¦π‘–π‘–π‘˜π‘˜βˆ’1𝑒𝑒

βˆ’π‘¦π‘¦π‘–π‘–πœƒπœƒ

kπœƒπœƒ ≑ πœ‡πœ‡; 1/πœ‡πœ‡ = 𝑋𝑋𝑋𝑋

GLM

Page 5: DATA ANALYSIS USING HIERARCHICAL GENERALIZED LINEAR MODELS

Gaussian

Poisson Binomial

Gamma

𝐿𝐿 = �𝑖𝑖=1

𝑛𝑛12πœ‹πœ‹πœŽπœŽ

𝑒𝑒12𝜎𝜎2 π‘¦π‘¦π‘–π‘–βˆ’πœ‡πœ‡ 2

πœ‡πœ‡ = 𝑋𝑋𝑋𝑋

𝐿𝐿 = �𝑖𝑖=1

π‘›π‘›πœ‡πœ‡π‘¦π‘¦π‘–π‘–π‘’π‘’βˆ’πœ‡πœ‡π‘¦π‘¦π‘–π‘–!

log(πœ‡πœ‡) = 𝑋𝑋𝑋𝑋

𝐿𝐿 = �𝑖𝑖=1

𝑛𝑛𝑛𝑛𝑦𝑦𝑖𝑖 πœ‡πœ‡π‘¦π‘¦π‘–π‘–(1 βˆ’ πœ‡πœ‡)π‘›π‘›βˆ’π‘¦π‘¦π‘–π‘–

𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙(πœ‡πœ‡) = 𝑋𝑋𝑋𝑋

𝐿𝐿 = �𝑖𝑖=1

𝑛𝑛1

Ξ“(π‘˜π‘˜) πœƒπœƒπ‘˜π‘˜π‘¦π‘¦π‘–π‘–π‘˜π‘˜βˆ’1𝑒𝑒

βˆ’π‘¦π‘¦π‘–π‘–πœƒπœƒ

kπœƒπœƒ ≑ πœ‡πœ‡; 1/πœ‡πœ‡ = 𝑋𝑋𝑋𝑋

GLMCommon estimation algorithm using iterativeregression - Fast and easy to implement- Linear regression model checking tools!!!

Page 6: DATA ANALYSIS USING HIERARCHICAL GENERALIZED LINEAR MODELS

Lee Y. & Nelder J. A. (1996) ”Hierarchicalgeneralized linear models” JRSS B 619-678

GLM approach for fittingβ€’ Linear mixed modelsβ€’ Generalized linear mixed models (Laplace approximation)

β€’ Mixed models with non-Gaussian random effectsβ€’ Above models + dispersion modelling

Page 7: DATA ANALYSIS USING HIERARCHICAL GENERALIZED LINEAR MODELS

Coming out July 2017

Page 8: DATA ANALYSIS USING HIERARCHICAL GENERALIZED LINEAR MODELS

Linear model

Generalized linear model (GLM)

Joint GLM Generalized linear model

including dispersion model with fixed effects

Linear mixed model (LMM)

Generalized linear mixedmodel (GLMM)

Generalized linear model including Gaussian random effects

Hierarchical GLM (HGLM)Generalized linear model including Gaussian

and/or non-Gaussian random effects. Dispersion can be modelled using fixed

effects.

Page 9: DATA ANALYSIS USING HIERARCHICAL GENERALIZED LINEAR MODELS

Linear model

Generalized linear model (GLM)

Joint GLM Generalized linear model

including dispersion model with fixed effects

Linear mixed model (LMM)

Generalized linear mixedmodel (GLMM)

Generalized linear model including Gaussian random effects

Double HGLM (DHGLM)HGLM including dispersion model with both fixed and random effects

Frailty HGLMHGLMs for survival analysis including

competing risk models

Structural Equation Models (SEM)

Hierarchical GLM (HGLM)Generalized linear model including Gaussian

and/or non-Gaussian random effects. Dispersion can be modelled using fixed

effects.

HGLMs with correlated random effects

Including spatial, temporal correlations, splines, GAM.

Factor analysis

Page 10: DATA ANALYSIS USING HIERARCHICAL GENERALIZED LINEAR MODELS

Why use HGLMs?

β€’ Fast deterministic algorithms for complex models

β€’ All parts of the model are checkableβ€’ Predictions of unobservables

Page 11: DATA ANALYSIS USING HIERARCHICAL GENERALIZED LINEAR MODELS

Crack growth data

Hudak et al. (1978) presented data from an experiment where crack lengths are measured on a compact tension steel test.

β€’ There are 21 metallic specimens with the crack lengths recorded every 104 cycles.

β€’ y = increment of crack lengthβ€’ Covariate for the mean part of the model: crack length

previously recorded (crack0)β€’ Covariate for the dispersion part of the model: number of cycles

(cycle)

Page 12: DATA ANALYSIS USING HIERARCHICAL GENERALIZED LINEAR MODELS
Page 13: DATA ANALYSIS USING HIERARCHICAL GENERALIZED LINEAR MODELS

## GLM ##res_glm <- glm(y ~ crack0, family= Gamma(link=log), data=data_crack_growth)

Page 14: DATA ANALYSIS USING HIERARCHICAL GENERALIZED LINEAR MODELS

## GLM ##res_glm <- glm(y ~ crack0, family= Gamma(link=log), data=data_crack_growth)

library(hglm)## GLMM ##res_glmm1 <- hglm2(y ~ crack0 + (1|specimen), family = Gamma(link=log) , data = data_crack_growth )

Page 15: DATA ANALYSIS USING HIERARCHICAL GENERALIZED LINEAR MODELS

## GLM ##res_glm <- glm(y ~ crack0, family= Gamma(link=log), data=data_crack_growth)

library(hglm)## GLMM ##res_glmm1 <- hglm2(y ~ crack0 + (1|specimen), family = Gamma(link=log) , data = data_crack_growth )

library(dhglm)## HGLM I ##model_mu <- DHGLMMODELING(Model="mean", Link="log",

LinPred = y ~ crack0 + (1|specimen), RandDist = "inverse-gamma")model_phi <- DHGLMMODELING(Model="dispersion")res_hglm1 <- dhglmfit(RespDist="gamma", DataMain=data_crack_growth,

MeanModel=model_mu, DispersionModel=model_phi)

Page 16: DATA ANALYSIS USING HIERARCHICAL GENERALIZED LINEAR MODELS

## GLM ##res_glm <- glm(y ~ crack0, family= Gamma(link=log), data=data_crack_growth)

library(hglm)## GLMM ##res_glmm1 <- hglm2(y ~ crack0 + (1|specimen), family = Gamma(link=log) , data = data_crack_growth )

library(dhglm)## HGLM I ##model_mu <- DHGLMMODELING(Model="mean", Link="log",

LinPred = y ~ crack0 + (1|specimen), RandDist = "inverse-gamma")model_phi <- DHGLMMODELING(Model="dispersion")res_hglm1 <- dhglmfit(RespDist="gamma", DataMain=data_crack_growth,

MeanModel=model_mu, DispersionModel=model_phi)## HGLM II ##model_mu <- DHGLMMODELING(Model="mean", Link="log",

LinPred = y ~ crack0 + (1|specimen), RandDist="inverse-gamma")model_phi <- DHGLMMODELING(Model = "dispersion", Link = "log",

LinPred = phi ~ cycle)res_hglm2 <- dhglmfit(RespDist = "gamma", DataMain = data_crack_growth,

MeanModel = model_mu, DispersionModel = model_phi)

V(y)= πœ™πœ™πœ‡πœ‡2log(πœ™πœ™)=Xb

Page 17: DATA ANALYSIS USING HIERARCHICAL GENERALIZED LINEAR MODELS

## GLM ##res_glm <- glm(y ~ crack0, family= Gamma(link=log), data=data_crack_growth)

library(hglm)## GLMM ##res_glmm1 <- hglm2(y ~ crack0 + (1|specimen), family = Gamma(link=log) , data = data_crack_growth )

library(dhglm)## HGLM I ##model_mu <- DHGLMMODELING(Model="mean", Link="log",

LinPred = y ~ crack0 + (1|specimen), RandDist = "inverse-gamma")model_phi <- DHGLMMODELING(Model="dispersion")res_hglm1 <- dhglmfit(RespDist="gamma", DataMain=data_crack_growth,

MeanModel=model_mu, DispersionModel=model_phi)## HGLM II ##model_mu <- DHGLMMODELING(Model="mean", Link="log",

LinPred = y ~ crack0 + (1|specimen), RandDist="inverse-gamma")model_phi <- DHGLMMODELING(Model = "dispersion", Link = "log",

LinPred = phi ~ cycle)res_hglm2 <- dhglmfit(RespDist = "gamma", DataMain = data_crack_growth,

MeanModel = model_mu, DispersionModel = model_phi)## DHGLM I ##model_mu <- DHGLMMODELING(Model="mean", Link="log",

LinPred = y ~ crack0 + (1|specimen), RandDist="inverse-gamma")model_phi <- DHGLMMODELING(Model="dispersion", Link="log",

LinPred = phi ~ cycle + (1|specimen), RandDist="gaussian")res_dhglm1 <- dhglmfit(RespDist = "gamma", DataMain = data_crack_growth,

MeanModel = model_mu, DispersionModel = model_phi)

Page 18: DATA ANALYSIS USING HIERARCHICAL GENERALIZED LINEAR MODELS

hglm1

Page 19: DATA ANALYSIS USING HIERARCHICAL GENERALIZED LINEAR MODELS

hglm2

Page 20: DATA ANALYSIS USING HIERARCHICAL GENERALIZED LINEAR MODELS

dhglm1

Page 21: DATA ANALYSIS USING HIERARCHICAL GENERALIZED LINEAR MODELS

Use HGLMs!

β€’ Fast deterministic algorithms for complex models

β€’ All parts of the model are checkableβ€’ Predictions of unobservables

Page 22: DATA ANALYSIS USING HIERARCHICAL GENERALIZED LINEAR MODELS

[email protected]

www.larsronnegard.se

Page 23: DATA ANALYSIS USING HIERARCHICAL GENERALIZED LINEAR MODELS

[email protected]

DATA ANALYSIS USINGHIERARCHICAL GENERALIZED

LINEAR MODELS WITH R

Lee, RΓΆnnegΓ₯rd, Noh

www.larsronnegard.se


Recommended

Applied Bayesian Inference, KSU, April 29, 2012 § / §❻ Hierarchical (Multi-Stage) Generalized Linear Models Robert J. Tempelman 1
Applied Bayesian Inference, KSU, April 29, 2012 § / §❻ Hierarchical (Multi-Stage) Generalized Linear Models Robert J. Tempelman 1 Documents
Categorical Data Analysis - download.e-bookshelf.deΒ Β· 4.1 Generalized Linear Model, 116 4.2 Generalized Linear Models for Binary Data, 120 4.3 Generalized Linear Models for Counts,
Categorical Data Analysis - download.e-bookshelf.deΒ Β· 4.1 Generalized Linear Model, 116 4.2 Generalized Linear Models for Binary Data, 120 4.3 Generalized Linear Models for Counts, Documents
Linear Hierarchical Models
Linear Hierarchical Models Documents
Atlas of Digital Pathology: A Generalized Hierarchical
Atlas of Digital Pathology: A Generalized Hierarchical Documents
Chapter 8 Hierarchical Models2012/11/06 Β Β· Chapter 8 Hierarchical Models In the (generalized) linear models we’ve looked at so far, we’ve assumed that the observa-tions are independent
Chapter 8 Hierarchical Models2012/11/06 Β Β· Chapter 8 Hierarchical Models In the (generalized) linear models we’ve looked at so far, we’ve assumed that the observa-tions are independent Documents
Generalized Linear Models and ExtensionsΒ Β· 2010. 12. 16.Β Β· Session 1 - Generalized linear models † Introduction † Motivating examples † History † Generalized linear models
Generalized Linear Models and ExtensionsΒ Β· 2010. 12. 16.Β Β· Session 1 - Generalized linear models † Introduction † Motivating examples † History † Generalized linear models Documents
Introduction to General and Generalized Linear Models - Generalized Linear Models ...hmad/GLM/slides/lect07.pdfΒ Β· 2010. 10. 28.Β Β· Introduction to General and Generalized Linear
Introduction to General and Generalized Linear Models - Generalized Linear Models ...hmad/GLM/slides/lect07.pdfΒ Β· 2010. 10. 28.Β Β· Introduction to General and Generalized Linear Documents
Hierarchical Summary ROC Analysis: A Frequentist-Bayesian ...Hierarchical Summary ROC(HSROC) Model 1 Generalized non-linear mixed model 2 Focused on inferences about the SROC curve,
Hierarchical Summary ROC Analysis: A Frequentist-Bayesian ...Hierarchical Summary ROC(HSROC) Model 1 Generalized non-linear mixed model 2 Focused on inferences about the SROC curve, Documents
Hierarchical linear models - jarad.me
Hierarchical linear models - jarad.me Documents
Introduction to Generalized Linear Mixed Models, Matrix …€¦ · Introduction to Generalized Linear Mixed Models and Mathematical Statistics Generalized Linear Mixed Models Lecture
Introduction to Generalized Linear Mixed Models, Matrix …€¦ · Introduction to Generalized Linear Mixed Models and Mathematical Statistics Generalized Linear Mixed Models Lecture Documents
Atlas of Digital Pathology: A Generalized Hierarchical ...openaccess.thecvf.com/content_CVPR_2019/papers/Hosseini_Atlas_of...Β Β· Atlas of Digital Pathology: A Generalized Hierarchical
Atlas of Digital Pathology: A Generalized Hierarchical ...openaccess.thecvf.com/content_CVPR_2019/papers/Hosseini_Atlas_of...Β Β· Atlas of Digital Pathology: A Generalized Hierarchical Documents
Hierarchical Generalized Linear Models: The R …2 HGLMMM: Hierarchical Generalized Linear Models in R the joint modelling of the mean and dispersion structure. This estimation technique
Hierarchical Generalized Linear Models: The R …2 HGLMMM: Hierarchical Generalized Linear Models in R the joint modelling of the mean and dispersion structure. This estimation technique Documents
Generalized Linear Programming
Generalized Linear Programming Documents
GENERALIZED LINEAR MODELS IN BIOMEDICAL …€¦ · GENERALIZED LINEAR MODELS IN BIOMEDICAL STATISTICAL APPLICATIONS by ... GENERALIZED LINEAR MODELS IN BIOMEDICAL ... life-testingproblems,
GENERALIZED LINEAR MODELS IN BIOMEDICAL …€¦ · GENERALIZED LINEAR MODELS IN BIOMEDICAL STATISTICAL APPLICATIONS by ... GENERALIZED LINEAR MODELS IN BIOMEDICAL ... life-testingproblems, Documents
Prediction and calibration in generalized linear modelsΒ Β· PREDICTION AND CALIBRATION IN GENERALIZED LINEAR MODELS* ... Prediction and calibration in generalized linear models
Prediction and calibration in generalized linear modelsΒ Β· PREDICTION AND CALIBRATION IN GENERALIZED LINEAR MODELS* ... Prediction and calibration in generalized linear models Documents
Generalized Linear Models Introduction - Mark Irwinmarkirwin.net/stat135/Lecture/Lecture18.pdfΒ Β· Generalized Linear Models Introduction ... In the analysis of Generalized Linear
Generalized Linear Models Introduction - Mark Irwinmarkirwin.net/stat135/Lecture/Lecture18.pdfΒ Β· Generalized Linear Models Introduction ... In the analysis of Generalized Linear Documents
Introduction to Generalized Linear Models: Ordinal ...classroom.takasila.org/classroom/dataupload/...Β Β· The Generalized Linear Model DEFINITION: A Generalized Linear Model is a model
Introduction to Generalized Linear Models: Ordinal ...classroom.takasila.org/classroom/dataupload/...Β Β· The Generalized Linear Model DEFINITION: A Generalized Linear Model is a model Documents
Beyond the Generalized Linear Mixed Model: a Hierarchical Bayesian Perspective
Beyond the Generalized Linear Mixed Model: a Hierarchical Bayesian Perspective Documents