Particle/Astro-Machine Learning Seminar

An unfolding method based on conditional Invertible Neural Networks (cINN) using iterative training

Abstract: The unfolding of detector effects is crucial for the comparison of data to theory predictions. While traditional methods are limited to representing the data in a low number of dimensions, machine learning has enabled new unfolding techniques while retaining the full dimensionality. Generative networks like invertible neural networks (INN) enable a probabilistic unfolding, which maps individual data events to their corresponding unfolded probability distribution.

Read more about An unfolding method based on conditional Invertible Neural Networks (cINN) using iterative training

New Physics Learning Machine: an uncertainties-aware goodness-of-fit test for statistical anomalies detection

Abstract:

Read more about New Physics Learning Machine: an uncertainties-aware goodness-of-fit test for statistical anomalies detection

Finding the unknown with resonant anomaly detection

Abstract:

Resonant anomaly detection is an emerging class of unsupervised machine learning methods aiming to enhance the sensitivity of beyond-the-standard-model searches in high-energy physics. Contrary to conventional searches, where an explicit new physics signal model is targeted, this data-driven approach can cover a wide range of signals in a single analysis and with relatively minimal assumptions. I will provide an introduction to the concept and discuss a curated set of recent advances and challenges, as well as an ongoing real-world application.