Skip to yearly menu bar Skip to main content


Poster
in
Workshop: Structured Probabilistic Inference and Generative Modeling

Learning high-dimensional mixed models via amortized variational inference

Priscilla Ong · Manuel Haussmann · Harri Lähdesmäki

Keywords: [ Longitudinal Data ] [ latent variable model ] [ Linear Mixed Model ] [ Gaussian Processes ]


Abstract:

Modelling longitudinal data is an important yet challenging task. These datasets can be high-dimensional, consist of non-linear effects, and contain time-varying covariates. In this work, we leverage linear mixed models (LMMs) and amortized variational inference to provide conditional priors for VAEs, and propose LMM-VAE, a model that is scalable, interpretable, and shares theoretical connections to the GP-based VAEs. We empirically demonstrate that LMM-VAE performs competitively compared to existing approaches.

Chat is not available.