| |
| |
| |
Introduction: basic notions about Bayesian inference | |
| |
| |
| |
Basic notions | |
| |
| |
| |
Simple dependence structures | |
| |
| |
| |
Synthesis of conditional distributions | |
| |
| |
| |
Choice of the prior distribution | |
| |
| |
| |
Bayesian inference in the linear regression model | |
| |
| |
| |
Markov chain Monte Carlo methods | |
| |
| |
| |
Gibbs sampler | |
| |
| |
| |
Metropolis-Hastings algorithm | |
| |
| |
| |
Adaptive rejection Metropolis sampling | |
| |
| |
Problems | |
| |
| |
| |
Dynamic linear models | |
| |
| |
| |
Introduction | |
| |
| |
| |
A simple example | |
| |
| |
| |
State space models | |
| |
| |
| |
Dynamic linear models | |
| |
| |
| |
Dynamic linear models in package dlm | |
| |
| |
| |
Examples of nonlinear and non-Gaussian state space models | |
| |
| |
| |
State estimation and forecasting | |
| |
| |
| |
Filtering | |
| |
| |
| |
Kalman filter for dynamic linear models | |
| |
| |
| |
Filtering with missing observations | |
| |
| |
| |
Smoothing | |
| |
| |
| |
Forecasting | |
| |
| |
| |
The innovation process and model checking | |
| |
| |
| |
Controllability and observability of time-invariant DLMs | |
| |
| |
| |
Filter stability | |
| |
| |
Problems | |
| |
| |
| |
Model specification | |
| |
| |
| |
Classical tools for time series analysis | |
| |
| |
| |
Empirical methods | |
| |
| |
| |
ARIMA models | |
| |
| |
| |
Univariate DLMs for time series analysis | |
| |
| |
| |
Trend models | |
| |
| |
| |
Seasonal factor models | |
| |
| |
| |
Fourier form seasonal models | |
| |
| |
| |
General periodic components | |
| |
| |
| |
DLM representation of ARIMA models | |
| |
| |
| |
Example: estimating the output gap | |
| |
| |
| |
Regression models | |
| |
| |
| |
Models for multivariate time series | |
| |
| |
| |
DLMs for longitudinal data | |
| |
| |
| |
Seemingly unrelated time series equations | |
| |
| |
| |
Seemingly unrelated regression models | |
| |
| |
| |
Hierarchical DLMs | |
| |
| |
| |
Dynamic regression | |
| |
| |
| |
Common factors | |
| |
| |
| |
Multivariate ARMA models | |
| |
| |
Problems | |
| |
| |
| |
Models with unknown parameters | |
| |
| |
| |
Maximum likelihood estimation | |
| |
| |
| |
Bayesian inference | |
| |
| |
| |
Conjugate Bayesian inference | |
| |
| |
| |
Unknown covariance matrices: conjugate inference | |
| |
| |
| |
Specification of Wt by discount factors | |
| |
| |
| |
A discount factor model for time-varying Vt | |
| |
| |
| |
Simulation-based Bayesian inference | |
| |
| |
| |
Drawing the states given y1:T: forward filtering backward sampling | |
| |
| |
| |
General strategies for MCMC | |
| |
| |
| |
Illustration: Gibbs sampling for a local level model | |
| |
| |
| |
Unknown variances | |
| |
| |
| |
Constant unknown variances: d Inverse Gamma Prior | |
| |
| |
| |
Multivariate extensions | |
| |
| |
| |
A model for outliers and structural breaks | |
| |
| |
| |
Further examples | |
| |
| |
| |
Estimating the output gap: Bayesian inference | |
| |
| |
| |
Dynamic regression | |
| |
| |
| |
Factor models | |
| |
| |
Problems | |
| |
| |
| |
Sequential Monte Carlo methods | |
| |
| |
| |
The basic particle filter | |
| |
| |
| |
A simple example | |
| |
| |
| |
Auxiliary particle filter | |
| |
| |
| |
Sequential Monte Carlo with unknown parameters | |
| |
| |
| |
A simple example with unknown parameters | |
| |
| |
| |
Concluding remarks | |
| |
| |
| |
Useful distributions | |
| |
| |
| |
Matrix algebra: Singular Value Decomposition | |
| |
| |
Index | |
| |
| |
References | |