Track papers across labs, topics, and trends.
Search papers, labs, and topics across Lattice.
We introduce Amortized Structured Stochastic Variational Inference (ASSVI) to address the limitations of mean-field approximations in Gaussian Process Latent Variable Models (GPLVMs) for uncertainty estimation. By allowing conditional dependence between the variational posterior of latent variables and inducing points, ASSVI captures a richer representation of the data manifold's epistemic uncertainty. This enhanced variational inference significantly improves reconstruction metrics, demonstrating a more accurate modeling of the underlying data manifold.
Structured variational inference unlocks richer epistemic uncertainty estimation in GPLVMs, outperforming mean-field approximations on manifold reconstruction.
Many machine learning methods aim to approximate the lower-dimensional manifold on which the data lives. A desirable feature of such methods is that they should capture the epistemic uncertainty of this learned manifold. One model that achieves this is the Gaussian Process Latent Variable Model, in which a Gaussian Process (GP) mapping from the latent space provides an estimate of the uncertainty of the manifold. However, the effectiveness of this uncertainty estimation is limited by the mean-field variational approximation between the GP inducing points and the latent variables. In this work, we apply Amortized Structured Stochastic Variational Inference to allow the variational posterior for the latent space to be conditionally dependent on the value of the inducing points. We demonstrate that this more flexible variational posterior improves several metrics relating to the reconstruction of points on the data manifold.
TU Munich