| |
| |
Preface | |
| |
| |
| |
Review of Probability and Distribution Theory | |
| |
| |
| |
Probability and Random Variables | |
| |
| |
| |
Introduction | |
| |
| |
| |
Univariate Discrete Distributions | |
| |
| |
| |
The Bernoulli and Binomial Distributions | |
| |
| |
| |
The Poisson Distribution | |
| |
| |
| |
Binomial Distribution: Normal Approximation | |
| |
| |
| |
Univariate Continuous Distributions | |
| |
| |
| |
The Uniform, Beta, Gamma, Normal, and Student-t Distributions | |
| |
| |
| |
Multivariate Probability Distributions | |
| |
| |
| |
The Multinomial Distribution | |
| |
| |
| |
The Dirichlet Distribution | |
| |
| |
| |
The d-Dimensional Uniform Distribution | |
| |
| |
| |
The Multivariate Normal Distribution | |
| |
| |
| |
The Chi-square Distribution | |
| |
| |
| |
The Wishart and Inverse Wishart Distributions | |
| |
| |
| |
The Multivariate-t Distribution | |
| |
| |
| |
Distributions with Constrained Sample Space | |
| |
| |
| |
Iterated Expectations | |
| |
| |
| |
Functions of Random Variables | |
| |
| |
| |
Introduction | |
| |
| |
| |
Functions of a Single Random Variable | |
| |
| |
| |
Discrete Random Variables | |
| |
| |
| |
Continuous Random Variables | |
| |
| |
| |
Approximating the Mean and Variance | |
| |
| |
| |
Delta Method | |
| |
| |
| |
Functions of Several Random Variables | |
| |
| |
| |
Linear Transformations | |
| |
| |
| |
Approximating the Mean and Covariance Matrix | |
| |
| |
| |
Methods of Inference | |
| |
| |
| |
An Introduction to Likelihood Inference | |
| |
| |
| |
Introduction | |
| |
| |
| |
The Likelihood Function | |
| |
| |
| |
The Maximum Likelihood Estimator | |
| |
| |
| |
Likelihood Inference in a Gaussian Model | |
| |
| |
| |
Fisher's Information Measure | |
| |
| |
| |
Single Parameter Case | |
| |
| |
| |
Alternative Representation of Information | |
| |
| |
| |
Mean and Variance of the Score Function | |
| |
| |
| |
Multiparameter Case | |
| |
| |
| |
Cram�r-Rao Lower Bound | |
| |
| |
| |
Sufficiency | |
| |
| |
| |
Asymptotic Properties: Single Parameter Models | |
| |
| |
| |
Probability of the Data Given the Parameter | |
| |
| |
| |
Consistency | |
| |
| |
| |
Asymptotic Normality and Efficiency | |
| |
| |
| |
Asymptotic Properties: Multiparameter Models | |
| |
| |
| |
Functional Invariance | |
| |
| |
| |
Illustration of Functional Invariance | |
| |
| |
| |
Invariance in a Single Parameter Model | |
| |
| |
| |
Invariance in a Multiparameter Model | |
| |
| |
| |
Further Topics in Likelihood Inference | |
| |
| |
| |
Introduction | |
| |
| |
| |
Computation of Maximum Likelihood Estimates | |
| |
| |
| |
Evaluation of Hypotheses | |
| |
| |
| |
Likelihood Ratio Tests | |
| |
| |
| |
Confidence Regions | |
| |
| |
| |
Wald's Test | |
| |
| |
| |
Score Test | |
| |
| |
| |
Nuisance Parameters | |
| |
| |
| |
Loss of Efficiency Due to Nuisance Parameters | |
| |
| |
| |
Marginal Likelihoods | |
| |
| |
| |
Profile Likelihoods | |
| |
| |
| |
Analysis of a Multinomial Distribution | |
| |
| |
| |
Amount of Information per Observation | |
| |
| |
| |
Analysis of Linear Logistic Models | |
| |
| |
| |
The Logistic Distribution | |
| |
| |
| |
Likelihood Function under Bernoulli Sampling | |
| |
| |
| |
Mixed Effects Linear Logistic Model | |
| |
| |
| |
An Introduction to Bayesian Inference | |
| |
| |
| |
Introduction | |
| |
| |
| |
Bayes Theorem: Discrete Case | |
| |
| |
| |
Bayes Theorem: Continuous Case | |
| |
| |
| |
Posterior Distributions | |
| |
| |
| |
Bayesian Updating | |
| |
| |
| |
Features of Posterior Distributions | |
| |
| |
| |
Posterior Probabilities | |
| |
| |
| |
Posterior Quantiles | |
| |
| |
| |
Posterior Modes | |
| |
| |
| |
Posterior Mean Vector and Covariance Matrix | |
| |
| |
| |
Bayesian Analysis of Linear Models | |
| |
| |
| |
Introduction | |
| |
| |
| |
The Linear Regression Model | |
| |
| |
| |
Inference under Uniform Improper Priors | |
| |
| |
| |
Inference under Conjugate Priors | |
| |
| |
| |
Orthogonal Parameterization of the Model | |
| |
| |
| |
The Mixed Linear Model | |
| |
| |
| |
Bayesian View of the Mixed Effects Model | |
| |
| |
| |
Joint and Conditional Posterior Distributions | |
| |
| |
| |
Marginal Distribution of Variance Components | |
| |
| |
| |
Marginal Distribution of Location Parameters | |
| |
| |
| |
The Prior Distribution and Bayesian Analysis | |
| |
| |
| |
Introduction | |
| |
| |
| |
An Illustration of the Effect of Priors on Inferences | |
| |
| |
| |
A Rapid Tour of Bayesian Asymptotics | |
| |
| |
| |
Discrete Parameter | |
| |
| |
| |
Continuous Parameter | |
| |
| |
| |
Statistical Information and Entropy | |
| |
| |
| |
Information | |
| |
| |
| |
Entropy of a Discrete Distribution | |
| |
| |
| |
Entropy of a Joint and Conditional Distribution | |
| |
| |
| |
Entropy of a Continuous Distribution | |
| |
| |
| |
Information about a Parameter | |
| |
| |
| |
Fisher's Information Revisited | |
| |
| |
| |
Prior and Posterior Discrepancy | |
| |
| |
| |
Priors Conveying Little Information | |
| |
| |
| |
The Uniform Prior | |
| |
| |
| |
Other Vague Priors | |
| |
| |
| |
Maximum Entropy Prior Distributions | |
| |
| |
| |
Reference Prior Distributions | |
| |
| |
| |
Bayesian Assessment of Hypotheses and Models | |
| |
| |
| |
Introduction | |
| |
| |
| |
Bayes Factors | |
| |
| |
| |
Definition | |
| |
| |
| |
Interpretation | |
| |
| |
| |
The Bayes Factor and Hypothesis Testing | |
| |
| |
| |
Influence of the Prior Distribution | |
| |
| |
| |
Nested Models | |
| |
| |
| |
Approximations to the Bayes Factor | |
| |
| |
| |
Partial and Intrinsic Bayes Factors | |
| |
| |
| |
Estimating the Marginal Likelihood | |
| |
| |
| |
Goodness of Fit and Model Complexity | |
| |
| |
| |
Goodness of Fit and Predictive Ability of a Model | |
| |
| |
| |
Analysis of Residuals | |
| |
| |
| |
Predictive Ability and Predictive Cross-Validation | |
| |
| |
| |
Bayesian Model Averaging | |
| |
| |
| |
General | |
| |
| |
| |
Definitions | |
| |
| |
| |
Predictive Ability of BMA | |
| |
| |
| |
Approximate Inference Via the EM Algorithm | |
| |
| |
| |
Introduction | |
| |
| |
| |
Complete and Incomplete Data | |
| |
| |
| |
The EM Algorithm | |
| |
| |
| |
Form of the Algorithm | |
| |
| |
| |
Derivation | |
| |
| |
| |
Monotonic Increase of ln p (�y) | |
| |
| |
| |
The Missing Information Principle | |
| |
| |
| |
Complete, Observed and Missing Information | |
| |
| |
| |
Rate of Convergence of the EM Algorithm | |
| |
| |
| |
EM Theory for Exponential Families | |
| |
| |
| |
Standard Errors and Posterior Standard Deviations | |
| |
| |
| |
The Method of Louis | |
| |
| |
| |
Supplemented EM Algorithm (SEM) | |
| |
| |
| |
The Method of Oakes | |
| |
| |
| |
Examples | |
| |
| |
| |
Markov Chain Monte Carlo Methods | |
| |
| |
| |
An Overview of Discrete Markov Chains | |
| |
| |
| |
Introduction | |
| |
| |
| |
Definitions | |
| |
| |
| |
State of the System after n-Steps | |
| |
| |
| |
Long-Term Behavior of the Markov Chain | |
| |
| |
| |
Stationary Distribution | |
| |
| |
| |
Aperiodicity and Irreducibility | |
| |
| |
| |
Reversible Markov Chains | |
| |
| |
| |
Limiting Behavior | |
| |
| |
| |
Markov Chain Monte Carlo | |
| |
| |
| |
Introduction | |
| |
| |
| |
Preliminaries | |
| |
| |
| |
Notation | |
| |
| |
| |
Transition Kernels | |
| |
| |
| |
Varying Dimensionality | |
| |
| |
| |
An Overview of Markov Chain Monte Carlo | |
| |
| |
| |
The Metropolis-Hastings Algorithm | |
| |
| |
| |
An Informal Derivation | |
| |
| |
| |
A More Formal Derivation | |
| |
| |
| |
The Gibbs Sampler | |
| |
| |
| |
Fully Conditional Posterior Distributions | |
| |
| |
| |
The Gibbs Sampling Algorithm | |
| |
| |
| |
Langevin-Hastings Algorithm | |
| |
| |
| |
Reversible Jump MCMC | |
| |
| |
| |
The Invariant Distribution | |
| |
| |
| |
Generating the Proposal | |
| |
| |
| |
Specifying the Reversibility Condition | |
| |
| |
| |
Derivation of the Acceptance Probability | |
| |
| |
| |
Deterministic Proposals | |
| |
| |
| |
Generating Proposals via the Identity Mapping | |
| |
| |
| |
Data Augmentation | |
| |
| |
| |
Implementation and Analysis of MCMC Samples | |
| |
| |
| |
Introduction | |
| |
| |
| |
A Single Long Chain or Several Short Chains? | |
| |
| |
| |
Convergence Issues | |
| |
| |
| |
Effect of Posterior Correlation on Convergence | |
| |
| |
| |
Monitoring Convergence | |
| |
| |
| |
Inferences from the MCMC Output | |
| |
| |
| |
Estimators of Posterior Quantities | |
| |
| |
| |
Monte Carlo Variance | |
| |
| |
| |
Sensitivity Analysis | |
| |
| |
| |
Applications in Quantitative Genetics | |
| |
| |
| |
Gaussian and Thick-Tailed Linear Models | |
| |
| |
| |
Introduction | |
| |
| |
| |
The Univariate Linear Additive Genetic Model | |
| |
| |
| |
A Gibbs Sampling Algorithm | |
| |
| |
| |
Additive Genetic Model with Maternal Effects | |
| |
| |
| |
Fully Conditional Posterior Distributions | |
| |
| |
| |
The Multivariate Linear Additive Genetic Model | |
| |
| |
| |
Fully Conditional Posterior Distributions | |
| |
| |
| |
A Blocked Gibbs Sampler for Gaussian Linear Models | |
| |
| |
| |
Linear Models with Thick-Tailed Distributions | |
| |
| |
| |
Motivation | |
| |
| |
| |
A Student-t Mixed Effects Model | |
| |
| |
| |
Model with Clustered Random Effects | |
| |
| |
| |
Parameterizations and the Gibbs Sampler | |
| |
| |
| |
Threshold Models for Categorical Responses | |
| |
| |
| |
Introduction | |
| |
| |
| |
Analysis of a Single Polychotomous Trait | |
| |
| |
| |
Sampling Model | |
| |
| |
| |
Prior Distribution and Joint Posterior Density | |
| |
| |
| |
Fully Conditional Posterior Distributions | |
| |
| |
| |
The Gibbs Sampler | |
| |
| |
| |
Analysis of a Categorical and a Gaussian Trait | |
| |
| |
| |
Sampling Model | |
| |
| |
| |
Prior Distribution and Joint Posterior Density | |
| |
| |
| |
Fully Conditional Posterior Distributions | |
| |
| |
| |
The Gibbs Sampler | |
| |
| |
| |
Implementation with Binary Traits | |
| |
| |
| |
Bayesian Analysis of Longitudinal Data | |
| |
| |
| |
Introduction | |
| |
| |
| |
Hierarchical or Multistage Models | |
| |
| |
| |
First Stage | |
| |
| |
| |
Second Stage | |
| |
| |
| |
Third Stage | |
| |
| |
| |
Joint Posterior Distribution | |
| |
| |
| |
Two-Step Approximate Bayesian Analysis | |
| |
| |
| |
Estimating Location Parameters | |
| |
| |
| |
Estimating Dispersion Parameters | |
| |
| |
| |
Special Case: Linear First Stage | |
| |
| |
| |
Computation via Markov Chain Monte Carlo | |
| |
| |
| |
Fully Conditional Posterior Distributions | |
| |
| |
| |
Analysis with Thick-Tailed Distributions | |
| |
| |
| |
First- and Second-Stage Models | |
| |
| |
| |
Fully Conditional Posterior Distributions | |
| |
| |
| |
Segregation and Quantitative Trait Loci Analysis | |
| |
| |
| |
Introduction | |
| |
| |
| |
Segregation Analysis Models | |
| |
| |
| |
Notation and Model | |
| |
| |
| |
Fully Conditional Posterior Distributions | |
| |
| |
| |
Some Implementation Issues | |
| |
| |
| |
QTL Models | |
| |
| |
| |
Models with a Single QTL | |
| |
| |
| |
Models with an Arbitrary Number of QTL | |
| |
| |
References | |
| |
| |
List of Citations | |
| |
| |
Subject Index | |