Non-Gaussian Process Regression

by   Yaman Kındap, et al.

Standard GPs offer a flexible modelling tool for well-behaved processes. However, deviations from Gaussianity are expected to appear in real world datasets, with structural outliers and shocks routinely observed. In these cases GPs can fail to model uncertainty adequately and may over-smooth inferences. Here we extend the GP framework into a new class of time-changed GPs that allow for straightforward modelling of heavy-tailed non-Gaussian behaviours, while retaining a tractable conditional GP structure through an infinite mixture of non-homogeneous GPs representation. The conditional GP structure is obtained by conditioning the observations on a latent transformed input space and the random evolution of the latent transformation is modelled using a Lévy process which allows Bayesian inference in both the posterior predictive density and the latent transformation function. We present Markov chain Monte Carlo inference procedures for this model and demonstrate the potential benefits compared to a standard GP.


page 1

page 2

page 3

page 4


Efficient Bayesian Inference of Sigmoidal Gaussian Cox Processes

We present an approximate Bayesian inference approach for estimating the...

Recovering Latent Signals from a Mixture of Measurements using a Gaussian Process Prior

In sensing applications, sensors cannot always measure the latent quanti...

Transport Gaussian Processes for Regression

Gaussian process (GP) priors are non-parametric generative models with a...

Multiplicative Latent Force Models

Bayesian modelling of dynamic systems must achieve a compromise between ...

Gaussian Process Regression with Heteroscedastic or Non-Gaussian Residuals

Gaussian Process (GP) regression models typically assume that residuals ...

Fully-probabilistic Terrain Modelling with Stochastic Variational Gaussian Process Maps

Gaussian processes (GPs) are becoming a standard tool to build terrain r...

Efficient Deep Gaussian Process Models for Variable-Sized Input

Deep Gaussian processes (DGP) have appealing Bayesian properties, can ha...

Please sign up or login with your details

Forgot password? Click here to reset