Dynamic shrinkage processes

DOI10.1111/RSSB.12325zbMATH Open1428.62397arXiv1707.00763OpenAlexW3101281330MaRDI QIDQ5204364FDOQ5204364

Authors: Daniel R. Kowal, David S. Matteson, David Ruppert 

Publication date: 4 December 2019

Published in: Journal of the Royal Statistical Society Series B: Statistical Methodology (Search for Journal in Brave)

Abstract: We propose a novel class of dynamic shrinkage processes for Bayesian time series and regression analysis. Building upon a global-local framework of prior construction, in which continuous scale mixtures of Gaussian distributions are employed for both desirable shrinkage properties and computational tractability, we model dependence among the local scale parameters. The resulting processes inherit the desirable shrinkage behavior of popular global-local priors, such as the horseshoe prior, but provide additional localized adaptivity, which is important for modeling time series data or regression functions with local features. We construct a computationally efficient Gibbs sampling algorithm based on a P'olya-Gamma scale mixture representation of the proposed process. Using dynamic shrinkage processes, we develop a Bayesian trend filtering model that produces more accurate estimates and tighter posterior credible intervals than competing methods, and apply the model for irregular curve-fitting of minute-by-minute Twitter CPU usage data. In addition, we develop an adaptive time-varying parameter regression model to assess the efficacy of the Fama-French five-factor asset pricing model with momentum added as a sixth factor. Our dynamic analysis of manufacturing and healthcare industry data shows that with the exception of the market risk, no other risk factors are significant except for brief periods.

Full work available at URL: https://arxiv.org/abs/1707.00763

Recommendations

• Dynamic variable selection with spike-and-slab process priors
• Achieving shrinkage in a time-varying parameter model framework
• Time-varying sparsity in dynamic regression models
• scientific article; zbMATH DE number 472962
• Local shrinkage rules, Lévy processes and regularized regression

zbMATH Keywords

regressiontime seriestrend filteringstochastic volatilityasset pricingdynamic linear model

Mathematics Subject Classification ID

Time series, auto-correlation, regression, etc. in statistics (GARCH) (62M10) Ridge regression; shrinkage estimators (Lasso) (62J07) Inference from stochastic processes and prediction (62M20) Applications of statistics to actuarial sciences and financial mathematics (62P05)

Cited In (25)

• Modeling tail risks of inflation using unobserved component quantile regressions
• Horseshoe Regularisation for Machine Learning in Complex and Deep Models¹
• Bayesian Change Point Detection with Spike-and-Slab Priors
• Parsimony inducing priors for large scale state-space models
• Detecting Changes in Covariance via Random Matrix Theory
• Bayesian Trend Filtering via Proximal Markov Chain Monte Carlo
• Horseshoe shrinkage methods for Bayesian fusion estimation
• Locally adaptive Bayesian isotonic regression using half shrinkage priors
• Modeling basal body temperature data using horseshoe process regression
• A unified view on Bayesian varying coefficient models
• Lasso meets horseshoe: a survey
• Bayesian Function-on-Scalars Regression for High-Dimensional Data
• Dynamic regression models for time-ordered functional data
• Dynamic variable selection with spike-and-slab process priors
• Fast and locally adaptive Bayesian quantile smoothing using calibrated variational approximations
• BAYESIAN DYNAMIC VARIABLE SELECTION IN HIGH DIMENSIONS
• Shrinkage strategies in some multiple multi-factor dynamical systems
• Dynamic shrinkage priors for large time-varying parameter regressions using scalable Markov chain Monte Carlo methods
• Bayesian Approaches to Shrinkage and Sparse Estimation
• Title not available (Why is that?)
• Social distancing merely stabilized COVID-19 in the United States
• Shrinkage with shrunken shoulders: Gibbs sampling shrinkage model posteriors with guaranteed convergence rates
• Time-dependent shrinkage of time-varying parameter regression models
• A Bayesian model of dose-response for cancer drug studies
• High-Dimensional Macroeconomic Forecasting Using Message Passing Algorithms

This page was built for publication: Dynamic shrinkage processes

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q5204364)