Fast Variational Learning in State-Space Gaussian Process Models

TL;DR

This paper introduces a novel, efficient variational inference method for state-space Gaussian process models that achieves linear-time computation, stability, and scalability to large datasets using Kalman recursions and JAX implementation.

Contribution

It extends conjugate-computation variational inference to enable fast, stable, and scalable inference in non-Gaussian state-space GP models with a new JAX-based implementation.

Findings

Achieves linear-time inference in state-space GP models.

Provides a numerically stable variational inference method.

Scales to time series with millions of data points.

Abstract

Gaussian process (GP) regression with 1D inputs can often be performed in linear time via a stochastic differential equation formulation. However, for non-Gaussian likelihoods, this requires application of approximate inference methods which can make the implementation difficult, e.g., expectation propagation can be numerically unstable and variational inference can be computationally inefficient. In this paper, we propose a new method that removes such difficulties. Building upon an existing method called conjugate-computation variational inference, our approach enables linear-time inference via Kalman recursions while avoiding numerical instabilities and convergence issues. We provide an efficient JAX implementation which exploits just-in-time compilation and allows for fast automatic differentiation through large for-loops. Overall, our approach leads to fast and stable variational inference in state-space GP models that can be scaled to time series with millions of data points.