Parameter uncertainties in weighted unbinned maximum likelihood fits

Christoph Langenbruch (I. Physikalisches Institut B, RWTH Aachen, Sommerfeldstr. 14, Aachen, 52074, Germany)

Parameter estimation via unbinned maximum likelihood fits is central for many analyses performed in high energy physics. Unbinned maximum likelihood fits using event weights, for example to statistically subtract background contributions via the sPlot formalism, or to correct for acceptance effects, have recently seen increasing use in the community. However, it is well known that the naive approach to the estimation of parameter uncertainties via the second derivative of the logarithmic likelihood does not yield confidence intervals with the correct coverage in the presence of event weights. This paper derives the asymptotically correct expressions and compares them with several commonly used approaches for the determination of parameter uncertainties, some of which are shown to not generally be asymptotically correct. In addition, the effect of uncertainties on event weights is discussed, including uncertainties that can arise from the presence of nuisance parameters in the determination of sWeights.

{
  "_oai": {
    "updated": "2022-07-08T06:33:00Z", 
    "id": "oai:repo.scoap3.org:69670", 
    "sets": [
      "EPJC"
    ]
  }, 
  "authors": [
    {
      "affiliations": [
        {
          "country": "Germany", 
          "value": "I. Physikalisches Institut B, RWTH Aachen, Sommerfeldstr. 14, Aachen, 52074, Germany", 
          "organization": "RWTH Aachen"
        }
      ], 
      "surname": "Langenbruch", 
      "email": "christoph.langenbruch@cern.ch", 
      "full_name": "Langenbruch, Christoph", 
      "given_names": "Christoph"
    }
  ], 
  "titles": [
    {
      "source": "Springer", 
      "title": "Parameter uncertainties in weighted unbinned maximum likelihood fits"
    }
  ], 
  "dois": [
    {
      "value": "10.1140/epjc/s10052-022-10254-8"
    }
  ], 
  "publication_info": [
    {
      "page_end": "32", 
      "journal_title": "European Physical Journal C", 
      "material": "article", 
      "journal_volume": "82", 
      "artid": "s10052-022-10254-8", 
      "year": 2022, 
      "page_start": "1", 
      "journal_issue": "5"
    }
  ], 
  "$schema": "http://repo.scoap3.org/schemas/hep.json", 
  "acquisition_source": {
    "date": "2022-07-08T06:30:57.061952", 
    "source": "Springer", 
    "method": "Springer", 
    "submission_number": "6476c590fe8711eca1a1661ec451a4e4"
  }, 
  "page_nr": [
    32
  ], 
  "license": [
    {
      "url": "https://creativecommons.org/licenses//by/4.0", 
      "license": "CC-BY-4.0"
    }
  ], 
  "copyright": [
    {
      "holder": "The Author(s)", 
      "year": "2022"
    }
  ], 
  "control_number": "69670", 
  "record_creation_date": "2022-05-18T12:00:59.623439", 
  "_files": [
    {
      "checksum": "md5:32f5f88619c95cc48033f244a9ec1c32", 
      "filetype": "xml", 
      "bucket": "6c5c8d25-99c4-46a6-a52c-4ea4abe9f514", 
      "version_id": "941b5ea0-3d70-4a75-8de5-5c09092689f3", 
      "key": "10.1140/epjc/s10052-022-10254-8.xml", 
      "size": 10459
    }, 
    {
      "checksum": "md5:19e0a0ca210ba8f8245643f073605cc7", 
      "filetype": "pdf/a", 
      "bucket": "6c5c8d25-99c4-46a6-a52c-4ea4abe9f514", 
      "version_id": "74e53f75-8f3f-4fa8-b2fe-f3b40608101e", 
      "key": "10.1140/epjc/s10052-022-10254-8_a.pdf", 
      "size": 1921835
    }
  ], 
  "collections": [
    {
      "primary": "European Physical Journal C"
    }
  ], 
  "abstracts": [
    {
      "source": "Springer", 
      "value": "Parameter estimation via unbinned maximum likelihood fits is central for many analyses performed in high energy physics. Unbinned maximum likelihood fits using event weights, for example to statistically subtract background contributions via the sPlot formalism, or to correct for acceptance effects, have recently seen increasing use in the community. However, it is well known that the naive approach to the estimation of parameter uncertainties via the second derivative of the logarithmic likelihood does not yield confidence intervals with the correct coverage in the presence of event weights. This paper derives the asymptotically correct expressions and compares them with several commonly used approaches for the determination of parameter uncertainties, some of which are shown to not generally be asymptotically correct. In addition, the effect of uncertainties on event weights is discussed, including uncertainties that can arise from the presence of nuisance parameters in the determination of sWeights."
    }
  ], 
  "imprints": [
    {
      "date": "2022-05-03", 
      "publisher": "Springer"
    }
  ]
}
Published on:
03 May 2022
Publisher:
Springer
Published in:
European Physical Journal C , Volume 82 (2022)
Issue 5
Pages 1-32
DOI:
https://doi.org/10.1140/epjc/s10052-022-10254-8
Copyrights:
The Author(s)
Licence:
CC-BY-4.0

Fulltext files: