Bayesian Inference of Vector Autoregressions with Tensor Decompositions
Bibliographic Data
| ID | 19417851 |
|---|---|
| Authors | Yiyong Luo (0000-0001-9448-9684, Department of Statistical Science, University College London, corresponding author), Jim E Griffin (0000-0002-4828-7368, Department of Statistical Science, University College London) |
| Year | 2025 |
| Volume | 43 |
| Issue | 4 |
| Pages | 941-955 |
| Publication date | 2025-10-02 |
| Open Access | Yes |
| Type | ARTICLE |
| Venue | Journal of Business and Economic Statistics (JOURNAL) |
| Journal identifiers | ISSN: 0735-0015 • E-ISSN: 1537-2707 |
| Publisher | Informa UK Limited (PUBLISHER • GB) |
| DOI | 10.1080/07350015.2024.2447302 |
| OpenAlex | W4406052319 |
| Language | EN |
| References cited | 42 |
Vector autoregression (VAR) is a popular model for analyzing multivariate economic time series. However, VARs can be over-parameterized if the numbers of variables and lags are moderately large. Tensor VAR, a recent solution to over-parameterization, treats the coefficient matrix as a third-order tensor and estimates the corresponding tensor decomposition to achieve parsimony. In this paper, we employ the Tensor VAR structure with a CANDECOMP/PARAFAC (CP) decomposition and use Bayesian inference to estimate parameters. Firstly, we determine the rank by imposing the Multiplicative Gamma Prior to the tensor margins, i.e. elements in the decomposition, and accelerate the computation with an adaptive inferential scheme. Secondly, to obtain interpretable margins, we propose an interweaving algorithm to improve the mixing of margins and identify the margins using a post-processing procedure. In an application to the US macroeconomic data, our models outperform standard VARs in point and density forecasting and yield a summary of the dynamic of the US economy
Bayesian inference · Bayesian probability · Econometrics · Inference · Pure mathematics · Computer Science · Gaussian Processes and Bayesian Inference · Mathematics · Statistical Methods and Inference · Tensor decomposition and applications · Applied Mathematics · Artificial Intelligence
Large Bayesian vector auto regressions
Drifts and volatilities
Forecasting and conditional projection using realistic prior distributions
Some Mathematical Notes on Three-Mode Factor Analysis
Macroeconomics and Reality
Inference from Iterative Simulation Using Multiple Sequences
Regression Shrinkage and Selection Via the Lasso
Bayesian Dynamic Tensor Regression
Large Order-Invariant Bayesian VARs with Stochastic Volatility
Adaptive Shrinkage in Bayesian Vector Autoregressive Models
Forecasting With Bayesian Vector Autoregressions—Five Years of Experience
Prior Selection for Vector Autoregressions
| Highly cited | No |
|---|