Gaussian Wasserstein Inference: Gaussian Measures meet Bayesian Deep Learning
Talk, Alan Turing Institute, Data-Centric Engineering Seminar, London, UK
The slides for a talk I gave about how we can use Gaussian Measures on the space of square-integrable functions to construct a highly flexible inference framework. We obtain state-of-the art results on benchmark data sets by combining deep neural networks with Gaussian measures in a novel way. The slides can be found here.