From Cost-Sensitive Classification to Tight F-measure Bounds

Kevin Bascol 1, 2 Rémi Emonet 1 Elisa Fromont 3 Amaury Habrard 1 Guillaume Metzler 1 Marc Sebban 1
3 LACODAM - Large Scale Collaborative Data Mining
Inria Rennes – Bretagne Atlantique , IRISA_D7 - GESTION DES DONNÉES ET DE LA CONNAISSANCE
Abstract : The F-measure is a classification performance measure, especially suited when dealing with imbalanced datasets, which provides a compromise between the precision and the recall of a classifier. As this measure is non convex and non linear, it is often indirectly optimized using cost-sensitive learning (that affects different costs to false positives and false negatives). In this article, we derive theoretical guarantees that give tight bounds on the best F-measure that can be obtained from cost-sensitive learning. We also give an original geometric interpretation of the bounds that serves as an inspiration for CONE, a new algorithm to optimize for the F-measure. Using 10 datasets exhibiting varied class imbalance, we illustrate that our bounds are much tighter than previous work and show that CONE learns models with either superior F-measures than existing methods or comparable but in fewer iterations.
Document type :
Conference papers
Complete list of metadatas

Cited literature [21 references]  Display  Hide  Download

https://hal.archives-ouvertes.fr/hal-02049763
Contributor : Guillaume Metzler <>
Submitted on : Tuesday, February 26, 2019 - 4:00:13 PM
Last modification on : Thursday, May 23, 2019 - 11:16:01 AM
Long-term archiving on : Monday, May 27, 2019 - 2:34:04 PM

File

aistats_2018.pdf
Files produced by the author(s)

Identifiers

  • HAL Id : hal-02049763, version 1

Citation

Kevin Bascol, Rémi Emonet, Elisa Fromont, Amaury Habrard, Guillaume Metzler, et al.. From Cost-Sensitive Classification to Tight F-measure Bounds. AISTATS 2019 - 22nd International Conference on Artificial Intelligence and Statistics, Apr 2019, Naha, Okinawa, Japan. pp.1245-1253. ⟨hal-02049763⟩

Share

Metrics

Record views

87

Files downloads

108