Sequential sampling of Gaussian process latent variable models
This addresses a computational bottleneck for researchers and practitioners using probabilistic models with sequential data, though it is incremental as it extends existing techniques.
The paper tackles the problem of scaling Gaussian process latent variable models with complex likelihoods to large sequential datasets by proposing a sequential sampling approximation for both latent variables and parameters, demonstrating strong performance in growing-data settings where non-sequential methods become unfeasible.
We consider the problem of inferring a latent function in a probabilistic model of data. When dependencies of the latent function are specified by a Gaussian process and the data likelihood is complex, efficient computation often involve Markov chain Monte Carlo sampling with limited applicability to large data sets. We extend some of these techniques to scale efficiently when the problem exhibits a sequential structure. We propose an approximation that enables sequential sampling of both latent variables and associated parameters. We demonstrate strong performance in growing-data settings that would otherwise be unfeasible with naive, non-sequential sampling.