Zobrazeno 1 - 6
of 6
pro vyhledávání: '"Zenn, Johannes"'
Predicting the physico-chemical properties of pure substances and mixtures is a central task in thermodynamics. Established prediction methods range from fully physics-based ab-initio calculations, which are only feasible for very simple systems, ove
Externí odkaz:
http://arxiv.org/abs/2406.08075
Autor:
Zenn, Johannes, Bamler, Robert
Differentiable annealed importance sampling (DAIS), proposed by Geffner & Domke (2021) and Zhang et al. (2021), allows optimizing over the initial distribution of AIS. In this paper, we show that, in the limit of many transitions, DAIS minimizes the
Externí odkaz:
http://arxiv.org/abs/2405.14840
The Street View House Numbers (SVHN) dataset is a popular benchmark dataset in deep learning. Originally designed for digit classification tasks, the SVHN dataset has been widely used as a benchmark for various other tasks including generative modeli
Externí odkaz:
http://arxiv.org/abs/2312.02168
Variational autoencoders (VAEs) are popular models for representation learning but their encoders are susceptible to overfitting (Cremer et al., 2018) because they are trained on a finite training set instead of the true (continuous) data distributio
Externí odkaz:
http://arxiv.org/abs/2310.19653
Autor:
Zenn, Johannes, Bamler, Robert
Annealed Importance Sampling (AIS) moves particles along a Markov chain from a tractable initial distribution to an intractable target distribution. The recently proposed Differentiable AIS (DAIS) (Geffner and Domke, 2021; Zhang et al., 2021) enables
Externí odkaz:
http://arxiv.org/abs/2304.14390
Autor:
Wenger, Jonathan, Krämer, Nicholas, Pförtner, Marvin, Schmidt, Jonathan, Bosch, Nathanael, Effenberger, Nina, Zenn, Johannes, Gessner, Alexandra, Karvonen, Toni, Briol, François-Xavier, Mahsereci, Maren, Hennig, Philipp
Probabilistic numerical methods (PNMs) solve numerical problems via probabilistic inference. They have been developed for linear algebra, optimization, integration and differential equation simulation. PNMs naturally incorporate prior information abo
Externí odkaz:
http://arxiv.org/abs/2112.02100