Veit D. Wild

Veit D. Wild

Research at the interface of Mathematics and Machine Learning

Publications

Generalized Variational Inference in Function Spaces: Gaussian Measures meet Bayesian Deep Learning

Published in arxiv, 2022

We develop a framework for generalized variational inference in infinite-dimensional function spaces and use it to construct a method termed Gaussian Wasserstein inference (GWI)…

Recommended citation: Veit D. Wild, Robert Hu and Dino Sejdinovic (2022). "Generalized Variational Inference in Function Spaces: Gaussian Measures meet Bayesian Deep Learning." arXiv preprint arXiv:2205.06342.

Variational Gaussian Processes: A Functional Analysis View

Published in International Conference on Artificial Intelligence and Statistics, 2022

Variational Gaussian process (GP) approximations have become a standard tool in fast GP inference. This technique requires a user to select variational features to increase efficiency.

Recommended citation: Veit D. Wild and George Wynne (2022). "Variational Gaussian Processes: A Functional Analysis View." International Conference on Artificial Intelligence and Statistics.

Bayesian Kernel Two-Sample Testing

Published in Journal of Computational and Graphical Statistics, 2022

In modern data analysis, nonparametric measures of discrepancies between random variables are particularly important. The subject is well-studied in the frequentist literature, while the development in the Bayesian setting is limited where applications are often restricted to univariate cases…

Recommended citation: Qinyi Zhang, Veit Wild, Sarah Filippi, Seth Flaxman and Dino Sejdinovic. "Generalized Variational Inference in Function Spaces: Gaussian Measures meet Bayesian Deep Learning." Journal of Computational and Graphical Statistics.

Connections and Equivalences between the Nyström Method and Sparse Variational Gaussian Processes

Published in arxiv, 2021

We investigate the connections between sparse approximation methods for making kernel methods and Gaussian processes (GPs) scalable to massive data, focusing on the Nyström method and Sparse Variational Gaussian Processes (SVGP).

Recommended citation: Veit Wild, Motonobu Kanagawa and Dino Sejdinovic (2010). "Connections and Equivalences between the Nyström Method and Sparse Variational Gaussian Processes." arXiv preprint arXiv:2106.01121 .