Source author record

Xuewen Yu

Xuewen Yu appears in the imported research catalog. Authorship, coauthor and topic links are available while profile ownership is still unclaimed.

ResearcherUnclaimed source record

econ.EM Artificial Intelligence math.ST Methodology Statistics Theory

Catalog footprint

What is connected

3works

5topics

4close collaborators

Actions

Connect this record

Open graph Browse works

Inspect adjacent papers, topics, institutions and collaborators without losing the researcher page.

Building this map preview

BZPEER is loading the nearby papers, people, topics and institutions for this page.

preprint2022arXiv

Fast and Accurate Variational Inference for Large Bayesian VARs with Stochastic Volatility

We propose a new variational approximation of the joint posterior distribution of the log-volatility in the context of large Bayesian VARs. In contrast to existing approaches that are based on local approximations, the new proposal provides a global approximation that takes into account the entire support of the joint distribution. In a Monte Carlo study we show that the new global approximation is over an order of magnitude more accurate than existing alternatives. We illustrate the proposed methodology with an application of a 96-variable VAR with stochastic volatility to measure global bank network connectedness.

preprint2022arXiv

Large Bayesian VARs with Factor Stochastic Volatility: Identification, Order Invariance and Structural Analysis

Vector autoregressions (VARs) with multivariate stochastic volatility are widely used for structural analysis. Often the structural model identified through economically meaningful restrictions--e.g., sign restrictions--is supposed to be independent of how the dependent variables are ordered. But since the reduced-form model is not order invariant, results from the structural analysis depend on the order of the variables. We consider a VAR based on the factor stochastic volatility that is constructed to be order invariant. We show that the presence of multivariate stochastic volatility allows for statistical identification of the model. We further prove that, with a suitable set of sign restrictions, the corresponding structural model is point-identified. An additional appeal of the proposed approach is that it can easily handle a large number of dependent variables as well as sign restrictions. We demonstrate the methodology through a structural analysis in which we use a 20-variable VAR with sign restrictions to identify 5 structural shocks.

preprint2020arXiv

Bayesian Learning of Causal Relationships for System Reliability

Causal theory is now widely developed with many applications to medicine and public health. However within the discipline of reliability, although causation is a key concept in this field, there has been much less theoretical attention. In this paper, we will demonstrate how some aspects of established causal methodology can be translated via trees, and more specifically chain event graphs, into domain of reliability theory to help the probability modeling of failures. We further show how various domain specific concepts of causality particular to reliability can be imported into more generic causal algebras and so demonstrate how these disciplines can inform each other. This paper is informed by a detailed analysis of maintenance records associated with a large electrical distribution company. Causal hypotheses embedded within these natural language texts are extracted and analyzed using the new graphical framework we introduced here.