Heterogeneous Multi-output Gaussian Process Prediction

Abstract

We present a novel extension of multi-output Gaussian processes for handling heterogeneous outputs. We assume that each output has its own likelihood function and use a vector-valued Gaussian process prior to jointly model the parameters in all likelihoods as latent functions. Our multi-output Gaussian process uses a covariance function with a linear model of coregionalisation form. Assuming conditional independence across the underlying latent functions together with an inducing variable framework, we are able to obtain tractable variational bounds amenable to stochastic variational inference. We illustrate the performance of the model on synthetic data and two real datasets: a human behavioral study and a demographic high-dimensional dataset.

References in corpus (3)

• Scalable Variational Gaussian Process Classification
• General Latent Feature Models for Heterogeneous Datasets
• A General Framework for Mixed Graphical Models

Cited by in corpus (8)

• Multioutput Gaussian Processes with Functional Data: A Study on Coastal Flood Hazard Assessment
• Variational Inference of Joint Models using Multivariate Gaussian Convolution Processes
• Regularized Multi-output Gaussian Convolution Process with Domain Adaptation
• Continual Multi-task Gaussian Processes
• Non-linear process convolutions for multi-output Gaussian processes
• Multi-task Learning for Aggregated Data using Gaussian Processes
• A Variational Autoencoder for Heterogeneous Temporal and Longitudinal Data
• Compositional Modeling of Nonlinear Dynamical Systems with ODE-based Random Features