keyboard_backspace go back
about lip

Funded Projects / details

Cofinanciado por:

FCT 
COMPETE2020 
Lisboa2020 
FEDER 

Project
BigDataHEP: Understanding Big Data in High Energy Physics: finding a needle in many haystacks

Code PTDC/FIS-PAR/29147/2017, POCI/01-0145-FEDER-029147

Beneficiary Entity

LIP - Laboratório de Instrumentação e Física Experimental de Partículas


Project summary

Despite of the tremendous success of the experimental confirmation of the Standard Model of Particle Physics, which culminated with the Higgs boson discovery by the Large Hadron Collider (LHC) experiments in 2012, there are still a significant number of unanswered questions. Important points, such as description of gravitational interactions in unified form with the other fundamental interactions, the fermion mass hierarchy, the number of fermion generations, the matter-antimatter asymmetry and the cosmological evidence for dark matter and dark energy, are not yet successfully explained. The large amount of data yet to be collected at the LHC, as well as the new generation of experiments attempting the direct detection of dark matter provide a unique opportunity to probe the current paradigm of Particle Physics and test hypotheses to go beyond it. Such purpose, however, requires the ability to fully exploit the available experimental data, since an efficient analysis might represent the difference between being able, or not, to observe subtle new physics effects or to probe rare properties of the matter.

The proposed project aims to develop complex analysis techniques to be used both in proton-proton and heavy ions collision data at the LHC and in the data collected by the LZ experiment. The goal is not to perform the data analysis itself, since this will be done by the experimental collaborations, but to develop new methods and tools to be used in such context, providing as well the computational tools required to efficiently process large datasets. Such data processing methods are particularly important in the context of the proposed machine learning techniques, since the training phase of these methods can be particularly challenging. Since some of the team members of this project are also active members of the ATLAS and LZ Collaborations the propagation of such tools to the experimental community is facilitated. Furthermore, this project also aims at using similar data analysis techniques in a different context, namely on the quality control of the industrial production of printed circuit boards. In this context machine learning techniques will be used to establish the source of the contaminants identified by chemical analysis. Some of the team members have previous experience in collaborating with the electronics industry, ensuring that the correct focus will be considered for the industrial application of the proposed methods. Finally, a key aspect of the project is the advanced training and knowledge dissemination in the field of data science.


Support under

Reforçar a investigação, o desenvolvimento tecnológico e a inovação

Region of Intervention

Centro  /  Lisboa  /  Norte

 

Funding

Total eligible cost

€ 239990

EU financial support

€ 177.725,09

National public financial support

€ 62.263,15



Publications


A Roadmap for HEP Software and Computing R&D for the 2020sArticle in international journal (with direct contribution from team)published
Finding new physics without learning about it: anomaly detection as a tool for searches at collidersArticle in international journal (with direct contribution from team)published
From the Bottom to the Top - Reconstruction of tt Events with Deep LearningArticle in international journal (with direct contribution from team)published
Machine Learning in High Energy Physics Community White PaperArticle in international journal (with direct contribution from team)published
Reconstruction of top quark pair dilepton decays in electron-positron collisionsArticle in international journal (with direct contribution from team)published
Search for large missing transverse momentum in association with one top-quark in proton-proton collisions at $sqrt{s}=13$ TeV with the ATLAS detectorArticle in international journal (with direct contribution from team)published
Search for pair and single production of vectorlike quarks in final states with at least one Z boson decaying into a pair of electrons or muons in pp collision data collected with the ATLAS detector at root s=13 TeVArticle in international journal (with direct contribution from team)published
Study of interference effects in the search for flavour-changing neutral current interactions involving the top quark and a photon or a Z boson at the LHCArticle in international journal (with direct contribution from team)published
Transferability of Deep Learning Models in Searches for New Physics at CollidersArticle in international journal (with direct contribution from team)published
Use of a Generalized Energy Mover's Distance in the Search for Rare Phenomena at CollidersArticle in international journal (with direct contribution from team)accepted

Presentations


Big data and machine learning at LIPOral presentation in international meeting
Big Data and Machine Learning in High Energy PhysicsPresentation in national conference
Data Science in High Energy PhysicsOral presentation in international conference
Deep Learning as a Tool for Generic Searches at CollidersOral presentation in international meeting
Deep Learning as a tool for Generic Searches at CollidersOral presentation in international conference
Do infinitamente pequeno ao infinitamente pequenoOutreach seminar
Física de Partículas: a ponte entre o infinitamente grande e o infinitamente pequenoOutreach seminar
Machine Learning in Chemistry: is a machine capable of outsmart a trained chemist?Seminar
Machine Learning in the Search for New Physics Phenomena at the LHCPoster presentation in national conference
Machine Learning na Física de Altas EnergiasPresentation in national conference
Measurements of Higgs boson production using decays to two b-quarks with the ATLAS detectorOral presentation in international conference
PlaCoR Plataforma para a Computação orientada ao RecursoPresentation in national conference
Probing the Standard Model and Beyond at the LHCOral presentation in advanced training events
Rare event detection in High Energy PhysicsOral presentation in international meeting
Searching for Rare Events at Colliders Using Deep LearningSeminar
The Portuguese participation in the upgrade of the Large Hadron Collider at CERNPresentation in national conference
Top-quark FCNC in production and decay processesOral presentation in international conference

Events


1st general meeting of the BigDataHEP projectCollaboration Meeting2019-01-11 / 2019-01-11
2nd general meeting of the BigDataHEP projectCollaboration Meeting2020-02-13 / 2020-02-13
Data Science in (Astro)Particle Physics and Cosmology: the Bridge to IndustryInternational Conference or Workshop2019-03-25 / 2019-03-29

Theses


Machine Learning in Analytical Chemistry: Applying Innovative Data Analysis Methods Using Chromatographic Techniques
New observables and techniques for the study of jets in hadron collisions
New physics phenomenology and data processing tools for the LZ experiment
PlaCor: Plataforma para a Computação Orientada ao Recurso
Search for Dark Matter in Monotop Events at the Large Hadron Collider
Search for FCNC in tZ trilepton events at the ATLAS experiment
Search for vector-like quarks in Zt/b+X events at ATLAS
Searching for dark matter with the ATLAS detector using unconventional signatures
Topic modelling for jets
Treino de redes neuronais profundas de forma distribuída

Team


Albano Agostinho Gomes Alves
Alexandre Miguel Ferreira Lindote
Ana Paula Pereira Peixoto
António Joaquim André Esteves
António Manuel da Silva Pina
Bruno Manuel Gonçalves Ribeiro
Bruno Miguel Leonardo Galhardo
Diogo Barros Gonçalves
Filipa Cavaco Reis Peres
Filipe Manuel Almeida Veloso
Francisco del Aguila Giménez
Henrique Manuel Peixoto Carvalho
João Pedro de Arruda Gonçalves
Johannes Erdmann
José Carlos Rufino Amaro
José Francisco Pimenta Fernandes
José Guilherme Teixeira de Almeida Milhano
José Santiago Perez
Juan Pedro Araque Espinosa
Kevin Kroeninger
Korinna Christine Zapp
Liliana Marisa Cunha Apolinário
Maria do Céu Neiva
Maura Gabriela Barros Teixeira
Miguel Castro Nunes Fiolhais
Miguel Correia dos Santos Crispim Romão
Nuno Filipe da Silva Fernandes de Castro
Paulo Alexandre Brinca Costa Brás
Pedro Miguel Martins Ferreira
Pier Parpot
Rui Alberto Serra Ribeiro dos Santos
Rute Costa Batalha Pedro
Tiago Dias do Vale
Tiago Fernandes Gonçalves
Tiago Nuno Fernandes Duarte
Tobias Golling
Vítor Serafim Pereira de Oliveira