Researchers release paper on the identifiable variational dynamic factor model iVDFM
Get up to speed
Minkey Chang and Jae-Young Kim have released a paper on the Identifiable Variational Dynamic Factor Model (iVDFM), a model designed for multivariate time series modeling with structural dynamics. It combines variational inference to provide partial identifiability guarantees within a known fuzzy class. The model's innovation term is conditioned on observed auxiliary context and mechanism embeddings, following a conditional exponential family prior. It is trained by maximizing the ELBO, enabling estimation of latent trajectories and predictive uncertainty. This research has been published in the ProbML 2026 conference proceedings, with the code open-sourced.
AI-generated from reports · 3 hr agoUpdated
Coverage timeline
Follow the coverage to understand different sides of the event.
- arXiv q-finResearch on Conditionally Identifiable Latent Representation Methods for Multivariate Time Series with Structural Dynamics
The study proposes the Identifiable Variational Dynamic Factor Model (iVDFM) for multivariate time series modeling. The model combines variational inference to provide partial identifiability guarantees within a known fuzzy class range. The model's innovation terms are conditioned on observed auxiliary context and mechanism embeddings, follow a conditional exponential family prior, and are trained by maximizing the ELBO to estimate uncertainty in latent trajectories and predictions. This research has been published in the ProbML 2026 conference proceedings, and the code is open source.
Trend of this event
Not enough consecutive observations yet; trend is not drawn.