no code implementations • 21 Jul 2021 • Théo Galy-Fajou, Manfred Opper
Gaussian Processes (\textbf{GPs}) are flexible non-parametric models with strong probabilistic interpretation.
no code implementations • pproximateinference AABI Symposium 2021 • Nikolai Zaki, Théo Galy-Fajou, Manfred Opper
Flow-based methods such as Stein Variational Gradient Descent caught a lot of interest due to their flexibility and the strong theory going with them.
no code implementations • pproximateinference AABI Symposium 2021 • Théo Galy-Fajou, Valerio Perrone, Manfred Opper
Bayesian inference is intractable for most practical problems and requires approximation schemes with several trade-offs.
1 code implementation • 26 Feb 2020 • Théo Galy-Fajou, Florian Wenzel, Manfred Opper
Building on the conjugate structure of the augmented model, we develop two inference methods.
3 code implementations • 23 May 2019 • Théo Galy-Fajou, Florian Wenzel, Christian Donner, Manfred Opper
We propose a new scalable multi-class Gaussian process classification approach building on a novel modified softmax likelihood function.