Variational Elliptical Processes
Abstract: We present elliptical processes—a family of non-parametric probabilistic models that subsumes the Gaussian processes and the Student's t processes. This generalization includes a range of new heavy-tailed behaviors while retaining computational tractability. The elliptical processes are based on a representation of elliptical distributions as a continuous mixture of Gaussian distributions. We parameterize this mixture distribution as a spline normalizing flow, which we train in two different ways using variational inference. The proposed form of the variational posterior enables a sparse variational elliptical process applicable to large-scale problems. We highlight some advantages compared to a Gaussian process through regression and classification experiments. Elliptical processes can replace Gaussian processes in several settings, including cases where the likelihood is non-Gaussian or when accurate tail modeling is essential.
Submission Length: Regular submission (no more than 12 pages of main content)
Assigned Action Editor: ~Sinead_Williamson1
Submission Number: 137
Loading