OSACA: Découverte d'attributs symboliques ordinaux

Abstract : This paper proposes to exploit heterogeneous data, i.e. data described by both numerical and categorical features, so as to discover whether, based on information provided by the numerical attributes, some categorical attributes actually are ordinal ones. The proposed 3-step methodology OSACA, first extracts gradual patterns from the numerical attributes ; it then applies mathematical morphology tools to induce an associated order on the categorical attributes. The third step evaluates the quality of the candidate rankings through measures derived from the rank entropy discrimination.
Complete list of metadatas

Cited literature [20 references]  Display  Hide  Download

https://hal.archives-ouvertes.fr/hal-01917965
Contributor : Christophe Marsala <>
Submitted on : Friday, October 25, 2019 - 11:20:03 AM
Last modification on : Monday, October 28, 2019 - 9:08:16 PM
Long-term archiving on: Sunday, January 26, 2020 - 2:04:31 PM

File

hal-01917965v1.pdf
Files produced by the author(s)

Identifiers

  • HAL Id : hal-01917965, version 1

Citation

Christophe Marsala, Anne Laurent, Marie-Jeanne Lesot, Maria Rifqi, Arnaud Castelltort. OSACA: Découverte d'attributs symboliques ordinaux. LFA: Logique Floue et ses Application, Nov 2018, Arras, France. pp.43-50. ⟨hal-01917965⟩

Share

Metrics

Record views

231

Files downloads

12