# Inference of cosmic-ray source properties by conditional invertible neural networks

@article{Bister2021InferenceOC, title={Inference of cosmic-ray source properties by conditional invertible neural networks}, author={Teresa Bister and Martin Erdmann and U. K{\"o}the and Josina Schulte}, journal={The European Physical Journal C}, year={2021}, volume={82} }

The inference of physical parameters from measured distributions constitutes a core task in physics data analyses. Among recent deep learning methods, so-called conditional invertible neural networks provide an elegant approach owing to their probability-preserving bijective mapping properties. They enable training the parameter-observation correspondence in one mapping direction and evaluating the parameter posterior distributions in the reverse direction. Here, we study the inference of…

## 3 Citations

### Two Invertible Networks for the Matrix Element Method

- Physics, Computer Science
- 2022

This work shows how a combination of two conditional generative neural networks encodes the QCD radiation and detector without any simplifying assumptions, and allows for likelihoods for individual events to be computed.

### What's Anomalous in LHC Jets?

- Physics
- 2022

Searches for anomalies are the main motivation for the LHC and define key analysis steps, including triggers. We discuss how LHC anomalies can be defined through probability density estimates,…

### Ultra-high-energy cosmic rays The intersection of the Cosmic and Energy Frontiers

- Physics, EducationAstroparticle Physics
- 2022

## References

SHOWING 1-10 OF 14 REFERENCES

### Stellar parameter determination from photometry using invertible neural networks

- Physics
- 2020

Photometric surveys with the Hubble Space Telescope (HST) remain one of the most efficient tools in astronomy to study stellar clusters with high resolution and deep coverage. Estimating physical…

### Inference from Iterative Simulation Using Multiple Sequences

- Computer Science
- 1992

The focus is on applied inference for Bayesian posterior distributions in real problems, which often tend toward normal- ity after transformations and marginalization, and the results are derived as normal-theory approximations to exact Bayesian inference, conditional on the observed simulations.

### Bayesian inversion for finite fault earthquake source models I—theory and algorithm

- Geology
- 2013

The estimation of finite fault earthquake source models is an inherently underdetermined
problem: there is no unique solution to the inverse problem of determining the rupture history
at depth as a…

### Data Analysis Recipes: Using Markov Chain Monte Carlo

- Mathematics
- 2017

It is argued that autocorrelation time is the most important test for convergence, as it directly connects to the uncertainty on the sampling estimate of any quantity of interest.

### Reinterpreting the development of extensive air showers initiated by nuclei and photons

- Physics
- 2013

Ultra-high energy cosmic rays (UHECRs) interacting with the atmosphere generate extensive air showers (EAS) of secondary particles. The depth corresponding to the maximum development of the shower,…

### CRPropa 3—a public astrophysical simulation framework for propagating extraterrestrial ultra-high energy particles

- PhysicsJournal of Cosmology and Astroparticle Physics
- 2016

We present the simulation framework CRPropa version 3 designed for efficient development of astrophysical predictions for ultra-high energy particles. Users can assemble modules of the most relevant…

### Combined fit of spectrum and composition data as measured by the Pierre Auger Observatory

- Physics
- 2015

We present a combined fit of a simple astrophysical model of UHECR sources to both the energy spectrum and mass composition data measured by the Pierre Auger Observatory. The fit has been performed…

### Probabilistic programming in Python using PyMC3

- Computer SciencePeerJ Comput. Sci.
- 2016

This paper is a tutorial-style introduction to PyMC3, a new open source Probabilistic Programming framework written in Python that uses Theano to compute gradients via automatic dierentiation as well as compile probabilistic programs on-the-fly to C for increased speed.

### The Pierre Auger Collaboration

- Physics
- 2013

The Pierre Auger Collaboration has reported evidence for anisotropies in the arrival directions of cosmic rays with energies larger thanEth = 55 EeV. There is a correlation above the isotropic…