Efficient estimation of conditionally linear and Gaussian state space models

Moura, Guilherme V.; Turatti, Douglas Eduardo

doi:10.1016/j.econlet.2014.07.019

Cited by 8 publications

(6 citation statements)

References 5 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Such a model would in particular be challenging for methods based on the Laplace approximation such as INLA (Rue et al, 2009), and recently, specialised computational methods for such models have been developed (see e.g. Moura and Turatti, 2014;Shephard, 2015;Li and Koopman, 2018). Here, on the other hand, it is shown that with a very modest coding effort under the proposed methodology, this model is easily tackled using the general purpose Stan software.…”

Section: Realistic Application -The Stock and Watson (2007) Modelmentioning

confidence: 99%

Dynamically Rescaled Hamiltonian Monte Carlo for Bayesian Hierarchical Models

Kleppe

2019

Journal of Computational and Graphical Statistics

View full text Add to dashboard Cite

Dynamically rescaled Hamiltonian Monte Carlo (DRHMC) is introduced as a computationally fast and easily implemented method for performing full Bayesian analysis in hierarchical statistical models. The method relies on introducing a modified parameterisation so that the re-parameterised target distribution has close to constant scaling properties, and thus is easily sampled using standard (Euclidian metric) Hamiltonian Monte Carlo. Provided that the parameterisations of the conditional distributions specifying the hierarchical model are "constant information parameterisations" (CIP), the relation between the modified-and original parameterisation is bijective, explicitly computed and admit exploitation of sparsity in the numerical linear algebra involved. CIPs for a large catalogue of statistical models are presented, and from the catalogue, it is clear that many CIPs are currently routinely used in statistical computing. A relation between the proposed methodology and a class of explicitly integrated Riemann manifold Hamiltonian Monte Carlo methods is discussed. The methodology is illustrated on several example models, including a model for inflation rates with multiple levels of non-linearly dependent latent variables.Supplementary materials are available online. IntroductionThe modelling of dependent data is routinely carried out using Bayesian hierarchical models in a diverse range of fields. The application of non-linear/non-Gaussian hierarchical models requires numerical methods for computing posterior distributions, predictions and so on, and with the demand

show abstract

Section: Realistic Application -The Stock and Watson (2007) Modelmentioning

confidence: 99%

Dynamically Rescaled Hamiltonian Monte Carlo for Bayesian Hierarchical Models

Kleppe

2019

Journal of Computational and Graphical Statistics

View full text Add to dashboard Cite

show abstract

“…which assumes that the future oil price consists of components with a direct interpretation that cannot be observed. ese three models are commonly used and proved that they can generate relatively accurate linear regression prediction [18,[34][35][36][37][42][43][44][45]. Same as the full model (equation (1)), these three restricted model versions can also do hierarchical parameter shrinkage and decide which variable parameter varies with time.…”

Section: Empirical Models and Computation Processesmentioning

confidence: 99%

Forecasting Oil Price by Hierarchical Shrinkage in Dynamic Parameter Models

Liu

Wei

Liu

et al. 2020

Discrete Dynamics in Nature and Society

View full text Add to dashboard Cite

The aim of this paper is to forecast monthly crude oil price with a hierarchical shrinkage approach, which utilizes not only LASSO for predictor selection, but a hierarchical Bayesian method to determine whether constant coefficient (CC) or time-varying parameter (TVP) predictive regression should be employed in each out-of-sample forecasting step. This newly developed method has the advantages of both model shrinkage and automatic switch between CC and TVP forecasting models; thus, this may produce more accurate predictions of crude oil prices. The empirical results show that this hierarchical shrinkage model can outperform many commonly used forecasting benchmark methods, such as AR, unobserved components stochastic volatility (UCSV), and multivariate regression models in forecasting crude oil price on various forecasting horizons.

show abstract

“…In their analysis, Stock and Watson (2007) fix the parameters ω 2 h = ω 2 g = 0.2. In a recent study, Moura and Turatti (2014) estimate the two parameters while maintaining the assumption that ω 2 h = ω 2 g = ω 2 . Using inflation in the G7 countries, they find that the estimates of ω 2 are statistically different from the calibrated value of 0.2 for a few countries.…”

Section: Application 1: Time-varying Volatility In Inflationmentioning

confidence: 99%

Specification Tests for Time-Varying Parameter Models with Stochastic Volatility

Chan

2015

SSRN Journal

View full text Add to dashboard Cite

We propose an easy technique to test for time-variation in coefficients and volatilities. Specifically, by using a noncentered parameterization for state space models, we develop a method to directly calculate the relevant Bayes factor using the Savage-Dickey density ratio-thus avoiding the computation of the marginal likelihood altogether. The proposed methodology is illustrated via two empirical applications. In the first application we test for time-variation in the volatility of inflation in the G7 countries. The second application investigates if there is substantial time-variation in the NAIRU in the US.

show abstract

Efficient estimation of conditionally linear and Gaussian state space models

Cited by 8 publications

References 5 publications

Dynamically Rescaled Hamiltonian Monte Carlo for Bayesian Hierarchical Models

Dynamically Rescaled Hamiltonian Monte Carlo for Bayesian Hierarchical Models

Forecasting Oil Price by Hierarchical Shrinkage in Dynamic Parameter Models

Specification Tests for Time-Varying Parameter Models with Stochastic Volatility

Contact Info

Product

Resources

About