Dynamic variable selection with spike-and-slab process priors
From MaRDI portal
Abstract: We address the problem of dynamic variable selection in time series regression with unknown residual variances, where the set of active predictors is allowed to evolve over time. To capture time-varying variable selection uncertainty, we introduce new dynamic shrinkage priors for the time series of regression coefficients. These priors are characterized by two main ingredients: smooth parameter evolutions and intermittent zeroes for modeling predictive breaks. More formally, our proposed Dynamic Spike-and-Slab (DSS) priors are constructed as mixtures of two processes: a spike process for the irrelevant coefficients and a slab autoregressive process for the active coefficients. The mixing weights are themselves time-varying and depend on lagged values of the series. Our DSS priors are probabilistically coherent in the sense that their stationary distribution is fully known and characterized by spike-and-slab marginals. For posterior sampling over dynamic regression coefficients, model selection indicators as well as unknown dynamic residual variances, we propose a Dynamic SSVS algorithm based on forward-filtering and backward-sampling. To scale our method to large data sets, we develop a Dynamic EMVS algorithm for MAP smoothing. We demonstrate, through simulation and a topical macroeconomic dataset, that DSS priors are very effective at separating active and noisy coefficients. Our fast implementation significantly extends the reach of spike-and-slab methods to large time series data.
Recommendations
Cites work
- scientific article; zbMATH DE number 3854103 (Why is no real title available?)
- scientific article; zbMATH DE number 4159863 (Why is no real title available?)
- scientific article; zbMATH DE number 472973 (Why is no real title available?)
- scientific article; zbMATH DE number 1034042 (Why is no real title available?)
- scientific article; zbMATH DE number 795289 (Why is no real title available?)
- scientific article; zbMATH DE number 3390199 (Why is no real title available?)
- A Bayesian analysis of the unit root in real exchange rates
- A Gaussian Mixture Autoregressive Model for Univariate Time Series
- A general theory of concave regularization for high-dimensional sparse estimation problems
- Achieving shrinkage in a time-varying parameter model framework
- An Autoregressive Process for Beta Random Variables
- Bayesian Variable Selection in Linear Regression
- Bayesian compressed vector autoregressions
- Bayesian emulation for multi-step optimization in decision problems
- Bayesian estimation of sparse signals with a continuous spike-and-slab prior
- Bayesian forecasting and dynamic models.
- Bayesian model selection for beta autoregressive processes
- Bayesian stochastic search for VAR model restrictions
- Combining Minimax Shrinkage Estimators
- DATA AUGMENTATION AND DYNAMIC LINEAR MODELS
- Dirichlet-Laplace priors for optimal shrinkage
- Dynamic shrinkage processes
- EMVS: the EM approach to Bayesian variable selection
- Forecasting economic time series using targeted predictors
- Forecasting inflation using dynamic model averaging
- Forecasting using a large number of predictors: is Bayesian shrinkage a valid alternative to principal components?
- Hierarchical shrinkage in time-varying parameter models
- Maximum a posteriori sequence estimation using Monte Carlo particle filters
- Maximum likelihood estimation via the ECM algorithm: A general framework
- Minimax multiple shrinkage estimation
- Nearly unbiased variable selection under minimax concave penalty
- On a Mixture Autoregressive Model
- On a logistic mixture autoregressive model
- Optimal predictive model selection.
- Prediction Via Orthogonalized Model Mixing
- Regularization and Variable Selection Via the Elastic Net
- Regularization of Wavelet Approximations
- Sparse and stable Markowitz portfolios
- Sparsity and Smoothness Via the Fused Lasso
- Spike and slab variable selection: frequentist and Bayesian strategies
- Stochastic Volatility: Likelihood Inference and Comparison with ARCH Models
- Stochastic model specification search for Gaussian and partial non-Gaussian state space models
- The Bayesian Lasso
- The spike-and-slab LASSO
- Time Varying Structural Vector Autoregressions and Monetary Policy
- Time series. Modeling, computation, and inference.
- Time-varying sparsity in dynamic regression models
- Variable Selection via Nonconcave Penalized Likelihood and its Oracle Properties
Cited in
(19)- High-Dimensional Macroeconomic Forecasting Using Message Passing Algorithms
- Fast and Flexible Bayesian Inference in Time-varying Parameter Regression Models
- Parsimony inducing priors for large scale state-space models
- Dynamic functional variable selection for multimodal mHealth data
- Locally time-varying parameter regression
- Incorporating grouping information into Bayesian Gaussian graphical model selection
- Time-varying sparsity in dynamic regression models
- Dynamic graphical models with variable selection for effective connectivity
- Dynamic sparsity on dynamic regression models
- High-dimensional confounding adjustment using continuous Spike and Slab priors
- A Bayesian Partially Observable Online Change Detection Approach with Thompson Sampling
- Dynamic shrinkage processes
- Comment: Regularization via Bayesian Penalty Mixing
- BAYESIAN DYNAMIC VARIABLE SELECTION IN HIGH DIMENSIONS
- Bayesian Approaches to Shrinkage and Sparse Estimation
- Discussion of ``Bayesian forecasting of multivariate time series: scalability, structure uncertainty and decisions
- Drift versus Shift: Decoupling Trends and Changepoint Analysis
- Time-dependent shrinkage of time-varying parameter regression models
- Bayesian variable selection for matrix autoregressive models
This page was built for publication: Dynamic variable selection with spike-and-slab process priors
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2057381)
