NNPDF Collaboration

NNPDF Collaboration
Name	NNPDF Collaboration
Formation	2001
Type	Research collaboration
Headquarters	CERN
Fields	High-energy physics, Quantum chromodynamics

NNPDF Collaboration is an international research collaboration that produces parton distribution functions for use in high-energy physics phenomenology, collider experiments, and theoretical studies. It develops statistical and computational techniques to extract quark and gluon distributions from experimental data collected at facilities and experiments such as CERN, European Organization for Nuclear Research, DESY, HERA, Tevatron, Fermilab, LHC, ATLAS experiment, CMS experiment, and LHCb experiment. The collaboration integrates methods from statistics, computational science, and quantum chromodynamics to provide inputs used by phenomenologists, experimental collaborations, and global analysis groups.

History

The collaboration originated in the early 2000s with researchers associated with institutions such as University of Cambridge, University of Oxford, Universidad de Zaragoza, University of Milano, INFN, IHEP, University of Edinburgh, and Universidad Complutense de Madrid aiming to address limitations in traditional parton distribution fits exemplified by groups like CTEQ, MSTW, and HERAPDF. Early work built on experimental results from HERA I, ZEUS, H1 (experiment), and theoretical foundations from perturbative Quantum chromodynamics studies at CERN Theory Division and academic centers such as IHES and Perimeter Institute. Over successive releases, the collaboration has responded to new datasets from Tevatron (collider), CDF (Collider Detector at Fermilab), DØ (DZero), and the Large Hadron Collider, updating methodologies as exemplified by major particle physics milestones like the Higgs boson discovery and precision measurements of the W boson and top quark.

Methodology

The collaboration’s methodology combines experimental input from DIS experiments at HERA and collider measurements from LHC experiments such as ATLAS experiment and CMS experiment with theoretical calculations from fixed-order perturbative Quantum chromodynamics and resummation frameworks developed at institutions like SLAC National Accelerator Laboratory and Brookhaven National Laboratory. It employs neural network parametrizations inspired by machine learning research at University of Toronto and Massachusetts Institute of Technology, and statistical techniques influenced by work at Institute for Advanced Study and Stanford University. The group uses cross-validation and Monte Carlo replica techniques rooted in practices at CERN and Fermilab to estimate uncertainties, incorporating heavy-quark schemes devised with input from researchers at DESY and LPNHE. Perturbative inputs reference calculations by collaborations associated with NNLOJET, APFEL, HOPPET, and theoretical developments from JHEP and Physical Review Letters authors.

Key Releases and Results

Major public releases have included sets that incorporate next-to-leading-order and next-to-next-to-leading-order corrections, aligning with theoretical advances by groups at IPPP, Nikhef, INFN Sezione di Milano, and UCL. Notable results provided updated determinations of the proton’s gluon distribution informed by jet measurements from Tevatron (collider) and LHC runs reported by ATLAS experiment and CMS experiment, and constraints on sea-quark asymmetries leveraging Drell–Yan data from E866/NuSea and LHCb experiment. Releases have influenced precision predictions for processes studied at CERN such as Higgs boson production, Drell–Yan process, and top quark pair production, with comparisons to global fits by CTEQ-TEA, MMHT, and ABMP analyses.

Collaborations and Affiliations

The collaboration maintains close ties with experimental groups including ATLAS experiment, CMS experiment, LHCb experiment, H1 (experiment), and ZEUS (experiment), and works alongside theory consortia at CERN Theory Division, IPPP, Nikhef, and national laboratories such as Brookhaven National Laboratory and Fermilab. It engages with standards bodies and infrastructure projects like HEPData, LHAPDF, and initiatives coordinated by ICHEP and EPS-HEP workshop participants. Members hold affiliations across universities and institutes including University of Oxford, Università di Milano, Universidad de Granada, University of Edinburgh, Universidad Complutense de Madrid, and national research agencies such as INFN and CNRS.

Impact on Particle Physics and Applications

The collaboration’s PDF sets are widely used in experimental analyses performed by ATLAS experiment, CMS experiment, LHCb experiment, and reinterpretations by theory groups at CERN and SLAC National Accelerator Laboratory, affecting cross section predictions, uncertainty estimates, and search strategies for signals such as Higgs boson properties and beyond-Standard-Model signatures investigated in ICHEP and Moriond conferences. Its uncertainty quantification techniques have influenced statistical practices adopted in global fits by CTEQ-TEA and MMHT and impacted phenomenology related to parton shower tuning at facilities like CERN and Brookhaven National Laboratory.

Software and Data Accessibility

The collaboration distributes PDF sets through community repositories such as LHAPDF and archives datasets and fit information compatible with HEPData and analysis tools used by ATLAS experiment and CMS experiment. Computational workflows leverage software projects including APFEL, HOPPET, NNLOJET, and machine-learning libraries developed in academic environments like University of Cambridge and Massachusetts Institute of Technology. Documentation and code releases are coordinated with open-science platforms used by the particle physics community and software stewardship efforts at CERN.

Criticisms and Validation Studies

Critiques have centered on methodological choices compared with alternative approaches from CTEQ-TEA, MMHT, and ABMP groups, prompting cross-validation studies performed by collaborations at CERN Theory Division, IPPP, and academic groups at University of Oxford and University of Cambridge. Validation studies evaluate neural-network parametrization stability, dataset selection, treatment of correlated systematic uncertainties from experiments like H1 (experiment) and ZEUS (experiment), and heavy-quark scheme differences studied by researchers at DESY and INFN. Ongoing benchmarking exercises at conferences such as ICHEP and workshops like PDF4LHC continue to compare results and quantify systematic differences.

Category:Particle physics collaborations