Multi-objective Bayesian optimization (MOBO) maximizes $M$ expensive black-box objectives with the goal of approximating the Pareto set, whose objective vectors cannot be improved in one component without sacrificing another. A common quality measure is hypervolume, the volume in objective space dominated by these vect...
Thomas Wedenig, Paul Plantosar, Robert Peharz· 0 citations
This work performs inference using a message passing scheme directly on the Structural Causal Models by representing structural mechanisms and probability densities in a common finite-dimensional function space and expresses inference through numerical quadrature and linear algebra without explicitly constructing condi...
Bayesian optimization (BO) over discrete spaces is challenging due to missing gradients and the sparsity of valid solutions. Latent-space BO sidesteps both issues by fitting a surrogate in the latent space of a generative model, which however substantially complicates the direct incorporation of domain knowledge that i...
Tommaso Ancilli, Thomas Wedenig, Henry Moss et al.· 0 citations
Bayesian physics-informed neural networks (B-PINNs) are a popular framework for parameter and state inference from sparse or noisy observations. They are commonly formulated via a collider structure, in which physical and trajectory parameters are assumed to be a priori independent and become coupled through virtual li...
This cumulative habilitation thesis studies probabilistic circuits (PCs) as a powerful and tractable framework for reasoning and learning under uncertainty in artificial intelligence (AI). It first advocates for probability as a core language for AI, emphasizing its connections to logic and information theory; the conc...