Description
Identify rare new physics process through an anomaly detection technique based on deep neural network (Graph Neural Network architecture).
Material for the exercise i.e. datasets and examples have been copied to the leonardo cluster and are available at:
/leonardo/home/usertrain/a08trb55/anomalyDetection/LHCO
Project proposal: description of the problem
Given a (pre-processed) dataset from fast simulation of a generic HEP detector containing a large number of events from Standard Model background processes and a test dataset containing both background and new physics signal events, design and train an anomaly detection model for anomaly detection of the NP processes.
Input dataset
Preprocessed data fro LHC OLYMPIC benchmark dataset, provided as numpy arrays
Machine learning methods
Graph Neural Networks and Auto-Encoder architectures
Goal and FOM
ROC curves, AUC
Project proposal: general context
High Energy Physics NP searches, Graph Neural Networks, Anomaly Detection
