Large-scale variational Gaussian state-space models

arXiv:2403.01371v1 Announce Type: cross
Abstract: We introduce an amortized variational inference algorithm and structured variational approximation for state-space models with nonlinear dynamics driven by Gaussian noise. Importantly, the proposed framework allows for efficient evaluation of the ELBO and low-variance stochastic gradient estimates without resorting to diagonal Gaussian approximations by exploiting (i) the low-rank structure of Monte-Carlo approximations to marginalize the latent state through the dynamics (ii) an inference network that approximates the update step with low-rank precision matrix updates (iii) …

abstract algorithm approximation arxiv cs.lg dynamics evaluation framework gradient inference low noise scale space state stat.ml stochastic type variance