Speaker
Description
The continuous growth in model complexity in high-energy physics (HEP) collider experiments demands increasingly time-consuming model fits. We show first results on the application of conditional invertible networks (cINNs) to this challenge. Specifically, we construct and train a cINN to learn the mapping from signal strength modifiers to observables and its inverse. The resulting network infers the posterior distribution of the signal strength modifiers rapidly and for low computational cost. We present performance indicators of such a setup including the treatment of systematic uncertainties and highlight the features of cINNs estimating a signal strength for HEP-data on simulations.
References
Inference of cosmic-ray source properties by conditional invertible neural networks, EPJC 82 (2022) 171
Significance
Evaluating trained cINNs is ultrafast and gives remarkable precision in reproducing arbitrary probability distributions from Gaussian distributions. Thus cINNs can be used for inference (ref. below). As a new approach to inference in HEP we construct a cINN, include systematic uncertainties, and present benchmark results using a typical simulated Higgs-plus-backgrounds physics case.
| Experiment context, if any | Simulations of CMS Experiment |
|---|
