2-5 April 2019
Jardí Botànic de València
Europe/Zurich timezone

A hybrid deep learning approach to vertexing

3 Apr 2019, 09:30
Jardí Botànic de València

Jardí Botànic de València

Carrer Quart 80, 46008 València, Spain
Talk 1: Machine learning, algorithms and theoretical analysis


Henry Fredrick Schreiner (University of Cincinnati (US))


In the transition to Run 3 in 2021, LHCb will undergo a major luminosity upgrade, going from 1.1 to 5.6 expected visible Primary Vertices (PVs) per event, and will adopt a purely software trigger. This has fueled increased interest in alternative highly-parallel and GPU friendly algorithms for tracking and reconstruction. We will present a novel prototype algorithm for vertexing in the LHCb upgrade conditions.

We use a custom kernel to transform the sparse 3D space of hits and tracks into a dense 1D dataset, and then apply Deep Learning techniques to find PV locations. By training networks on our kernels using several Convolutional Neural Network layers, we have achieved better than 90% efficiency with no more than 0.2 False Positives (FPs) per event. Beyond its physics performance, this algorithm also provides a rich collection of possibilities for visualization and study of 1D convolutional networks. We will discuss the design, performance, and future potential areas of improvement and study, such as possible ways to recover the full 3D vertex information.

Primary authors

Rui Fang (University of Cincinnati (US)) Henry Fredrick Schreiner (University of Cincinnati (US)) Michael David Sokoloff (University of Cincinnati (US)) Mr Constantin Niko Weisser (Massachusetts Inst. of Technology (US)) J Michael Williams (Massachusetts Inst. of Technology (US))

Presentation Materials

Your browser is out of date!

Update your browser to view this website correctly. Update my browser now