Project Description
The PhD student in this project will develop an advanced RTA-based analysis to efficiently select decays involving di-electron vertices with low invariant mass. They will use these advances to search for light dark photons with the LHCb detector focusing on a range of extremely low coupling so far unexplored by other experiments. The upgraded LHCb detector, which is planned to start data-taking in 2022, will allow for the first time to search for this kind of signature. It is the novel real-time detector readout in Run-3 that gives the possibility to use real-time analysis to separate dark photon candidates from the overwhelming background online. This challenging task will depend critically on the performance that can be achieved on modern computing architectures. The PhD student in this project will be trained in the usage of highly parallel architectures for the LHCb online reconstruction and will profit from collaboration with experts in the IGFAE at University of Santiago de Compostela on the use of GPUs. Through an internship with VERIZON, the PhD student will be trained in ML models development based on streams of data, and in the Amazon Web Services infrastructure for massively parallel training of large-scale ML models. The acquired ML processing skills will feed back into the main research project, allowing the development of advanced ML algorithms to optimise the electron classification and make it fast enough to be run in real time. This work will benefit from collaboration with the Dortmund experts and PhD students. These advances will also allow the PhD student to develop a tool to efficiently retrieve online low momentum photons through their conversion in di-electron pairs, a challenging task that requires an RTA-based approach.
Host country: Germany
Host beneficiary: University of Heidelberg
PhD-awarding institution: University of Heidelberg
Experiment affiliation: LHCb
Planned collaborations: Dortmund, VERIZON, IGFAE at University of Santiago de Compostela
Secondments
- TU Dortmund, Germany
- VERIZON Connect research centre in Florence, Italy
- IGFAE at University of Santiago de Compostela, Spain