Avoiding biases in binned fits

Binned maximum likelihood fits are an attractive option when analysing large datasets, but require care when computing likelihoods of continuous PDFs in bins. For many years the widely used statistical modelling package evaluated probabilities at the bin centre, leading to significant biases for str...

Descripción completa

Detalles Bibliográficos
Autores principales: Gligorov, V.V., Hageboeck, S., Nanut, T., Sciandra, A., Tou, D.Y.
Lenguaje:eng
Publicado: 2021
Materias:
physics.data-an
Other Fields of Physics
Acceso en línea:https://dx.doi.org/10.1088/1748-0221/16/08/T08004
http://cds.cern.ch/record/2767500
_version_ 1780971307012194304
author Gligorov, V.V.
Hageboeck, S.
Nanut, T.
Sciandra, A.
Tou, D.Y.
author_facet Gligorov, V.V.
Hageboeck, S.
Nanut, T.
Sciandra, A.
Tou, D.Y.
author_sort Gligorov, V.V.
collection CERN
description Binned maximum likelihood fits are an attractive option when analysing large datasets, but require care when computing likelihoods of continuous PDFs in bins. For many years the widely used statistical modelling package evaluated probabilities at the bin centre, leading to significant biases for strongly curved probability density functions. We demonstrate the biases with real-world examples, and introduce a PDF class to that removes these biases. The physics and computation performance of this new class are discussed.
id cern-2767500
institution Organización Europea para la Investigación Nuclear
language eng
publishDate 2021
record_format invenio
spelling cern-27675002023-01-31T08:07:25Zdoi:10.1088/1748-0221/16/08/T08004http://cds.cern.ch/record/2767500engGligorov, V.V.Hageboeck, S.Nanut, T.Sciandra, A.Tou, D.Y.Avoiding biases in binned fitsphysics.data-anOther Fields of PhysicsBinned maximum likelihood fits are an attractive option when analysing large datasets, but require care when computing likelihoods of continuous PDFs in bins. For many years the widely used statistical modelling package evaluated probabilities at the bin centre, leading to significant biases for strongly curved probability density functions. We demonstrate the biases with real-world examples, and introduce a PDF class to that removes these biases. The physics and computation performance of this new class are discussed.Binned maximum likelihood fits are an attractive option when analysing large datasets, but require care when computing likelihoods of continuous PDFs in bins. For many years the widely used statistical modelling package RooFit evaluated probabilities at the bin centre, leading to significant biases for strongly curved probability density functions. We demonstrate the biases with real-world examples, and introduce a PDF class to RooFit that removes these biases. The physics and computation performance of this new class are discussed.arXiv:2104.13879oai:cds.cern.ch:27675002021-04-28
spellingShingle physics.data-an
Other Fields of Physics
Gligorov, V.V.
Hageboeck, S.
Nanut, T.
Sciandra, A.
Tou, D.Y.
Avoiding biases in binned fits
title Avoiding biases in binned fits
title_full Avoiding biases in binned fits
title_fullStr Avoiding biases in binned fits
title_full_unstemmed Avoiding biases in binned fits
title_short Avoiding biases in binned fits
title_sort avoiding biases in binned fits
topic physics.data-an
Other Fields of Physics
url https://dx.doi.org/10.1088/1748-0221/16/08/T08004
http://cds.cern.ch/record/2767500
work_keys_str_mv AT gligorovvv avoidingbiasesinbinnedfits
AT hageboecks avoidingbiasesinbinnedfits
AT nanutt avoidingbiasesinbinnedfits
AT sciandraa avoidingbiasesinbinnedfits
AT toudy avoidingbiasesinbinnedfits

Ejemplares similares