F*: an interpretable transformation of the F-measure

David J. Hand; Peter Christen; Nishadi Kirielle

doi:10.1007/s10994-021-05964-1

F*: an interpretable transformation of the F-measure

David J. Hand^*, Peter Christen^*, Nishadi Kirielle

^*Corresponding author for this work

Research output: Contribution to journal › Article › peer-review

123 Citations (Scopus)

Abstract

The F-measure, also known as the F1-score, is widely used to assess the performance of classification algorithms. However, some researchers find it lacking in intuitive interpretation, questioning the appropriateness of combining two aspects of performance as conceptually distinct as precision and recall, and also questioning whether the harmonic mean is the best way to combine them. To ease this concern, we describe a simple transformation of the F-measure, which we call F^∗ (F-star), which has an immediate practical interpretation.

Original language	English
Pages (from-to)	451-456
Number of pages	6
Journal	Machine Learning
Volume	110
Issue number	3
DOIs	https://doi.org/10.1007/s10994-021-05964-1
Publication status	Published - Mar 2021

Access to Document

10.1007/s10994-021-05964-1

Cite this

@article{d5a91ff7771f4f1f8d9f100900fbb2f4,

title = "F*: an interpretable transformation of the F-measure",

abstract = "The F-measure, also known as the F1-score, is widely used to assess the performance of classification algorithms. However, some researchers find it lacking in intuitive interpretation, questioning the appropriateness of combining two aspects of performance as conceptually distinct as precision and recall, and also questioning whether the harmonic mean is the best way to combine them. To ease this concern, we describe a simple transformation of the F-measure, which we call F∗ (F-star), which has an immediate practical interpretation.",

keywords = "Classification, Error rate, F1-score, Interpretability, Performance, Precision, Recall",

author = "Hand, {David J.} and Peter Christen and Nishadi Kirielle",

note = "Publisher Copyright: {\textcopyright} 2021, The Author(s).",

year = "2021",

month = mar,

doi = "10.1007/s10994-021-05964-1",

language = "English",

volume = "110",

pages = "451--456",

journal = "Machine Learning",

issn = "0885-6125",

publisher = "Springer Netherlands",

number = "3",

}

TY - JOUR

T1 - F*

T2 - an interpretable transformation of the F-measure

AU - Hand, David J.

AU - Christen, Peter

AU - Kirielle, Nishadi

PY - 2021/3

Y1 - 2021/3

N2 - The F-measure, also known as the F1-score, is widely used to assess the performance of classification algorithms. However, some researchers find it lacking in intuitive interpretation, questioning the appropriateness of combining two aspects of performance as conceptually distinct as precision and recall, and also questioning whether the harmonic mean is the best way to combine them. To ease this concern, we describe a simple transformation of the F-measure, which we call F∗ (F-star), which has an immediate practical interpretation.

AB - The F-measure, also known as the F1-score, is widely used to assess the performance of classification algorithms. However, some researchers find it lacking in intuitive interpretation, questioning the appropriateness of combining two aspects of performance as conceptually distinct as precision and recall, and also questioning whether the harmonic mean is the best way to combine them. To ease this concern, we describe a simple transformation of the F-measure, which we call F∗ (F-star), which has an immediate practical interpretation.

KW - Classification

KW - Error rate

KW - F1-score

KW - Interpretability

KW - Performance

KW - Precision

KW - Recall

UR - http://www.scopus.com/inward/record.url?scp=85102812154&partnerID=8YFLogxK

U2 - 10.1007/s10994-021-05964-1

DO - 10.1007/s10994-021-05964-1

M3 - Article

SN - 0885-6125

VL - 110

SP - 451

EP - 456

JO - Machine Learning

JF - Machine Learning

IS - 3

ER -

F*: an interpretable transformation of the F-measure

Abstract

Access to Document

Other files and links

Fingerprint

Cite this