Conditionally Identifiable Latent Representation for Multivariate Time Series with Structural Dynamics
This work addresses the need for interpretable and reliable latent representations in time series analysis, though it appears incremental as it builds on existing iVAE methods.
The paper tackled the problem of learning identifiable latent factors from multivariate time series by proposing the Identifiable Variational Dynamic Factor Model (iVDFM), which ensures factors are identifiable up to permutation and affine transformations, and demonstrated improved factor recovery on synthetic data and competitive forecasting on real-world benchmarks.
We propose the Identifiable Variational Dynamic Factor Model (iVDFM), which learns latent factors from multivariate time series with identifiability guarantees. By applying iVAE-style conditioning to the innovation process driving the dynamics rather than to the latent states, we show that factors are identifiable up to permutation and component-wise affine (or monotone invertible) transformations. Linear diagonal dynamics preserve this identifiability and admit scalable computation via companion-matrix and Krylov methods. We demonstrate improved factor recovery on synthetic data, stable intervention accuracy on synthetic SCMs, and competitive probabilistic forecasting on real-world benchmarks.