10 papers
Practical and Scalable Hamiltonian Monte Carlo Without the Metropolis Test
Jakob Robnik, Reuben Cohn-Gordon, Uroš Seljak
Hamiltonian Monte Carlo and underdamped Langevin Monte Carlo are leading methods for sampling from high-dimensional distributions with differentiable densities. Both rely on numeri…
Microcanonical Hamiltonian Monte Carlo
Jakob Robnik, G. Bruno De Luca, Eva Silverstein +1
We develop Microcanonical Hamiltonian Monte Carlo (MCHMC), a class of models which follow a fixed energy Hamiltonian dynamics, in contrast to Hamiltonian Monte Carlo (HMC), which f…
Can Microcanonical Langevin Dynamics Leverage Mini-Batch Gradient Noise?
Emanuel Sommer, Kangning Diao, Jakob Robnik +2
Scaling inference methods such as Markov chain Monte Carlo to high-dimensional models remains a central challenge in Bayesian deep learning. A promising recent proposal, microcanon…
Radiological mapping and uncertainty quantification by a fast Microcanonical Langevin Monte Carlo sampler
Lei Pan, Jaewon Lee, Brian J. Quiter +3
Radiological mapping plays a critical role in nuclear emergency response and environmental management activities. A radiation image, representing the spatial and intensity distribu…
Exoplanet transit search at the detection limit: detection and false alarm vetting pipeline
Jakob Robnik, Uroš Seljak, Jon M. Jenkins +1
One of the primary mission goals of the Kepler space telescope was to detect Earth-like terrestrial planets in the habitable zone around Sun-like stars. These planets are at the de…
Faster parallel MCMC: Metropolis adjustment is best served warm
Jakob Robnik, Uroš Seljak
Despite the enormous success of Hamiltonian Monte Carlo and related Markov Chain Monte Carlo (MCMC) methods, sampling often still represents the computational bottleneck in scienti…