3 papers
cs.LG2026
Forecasting Multiple Observables with SCROLL: Score-Trained Uncertainty for Stochastic Dynamics
Pavel Prochazka
Forecasting a stochastic dynamical system rarely means a single number: one wants several observables---future state, threshold event, regime label---each with its own likelihood.…
cs.LG2026
Beliefs Beyond Posteriors: Local-Consistency Optimisation for Bayesian Neural Networks
Pavel Prochazka
The standard training objectives of Bayesian deep learning are posterior-seeking: their optimum over the belief is the posterior of a fitted model, or its KL projection. We show th…
cs.LG2024
Convolutional Signal Propagation: A Simple Scalable Algorithm for Hypergraphs
Pavel Procházka, Marek DÄdiÄ, Lukáš Bajer
Last decade has seen the emergence of numerous methods for learning on graphs, particularly Graph Neural Networks (GNNs). These methods, however, are often not directly applicable…