Research Seminar on AI: Inference of cosmic-ray source properties by conditional invertible neural networks
Tuesday, May 03, 2022, 4:00pm
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, Josina Schulte and her team study the inference of cosmic-ray source properties from cosmic-ray observations on Earth using extensive astrophysical simulations. Josina Schulte and her team compare the performance of conditional invertible neural networks (cINNs) with the frequently used Markov Chain Monte Carlo (MCMC) method. While cINNs are trained to directly predict the parameters' posterior distributions, the MCMC method extracts the posterior distributions through a likelihood function that matches simulations with observations. Overall, Josina Schulte and her team find good agreement between the physics parameters derived by the two different methods. As a result of its computational efficiency, the cINN method allows for a swift assessment of inference quality.
Josina Schulte is a PhD student at the physics institute III A in the group of Professor Erdmann, working on astroparticle physics with the Pierre Auger Observatory. Her research focuses are the search for possible source candidates of ultra-high-energy cosmic rays, and the application and study of modern analysis techniques, including methods based on Deep Learning.