Skip to Main content Skip to Navigation
New interface
Conference papers

Intégration incrémentale de contraintes pour le clustering avec la programmation par contraintes

Abstract : Actively involving an expert in a constrained clustering loop translates into an incremental process where expert constraints are added on the fly. However, intermediate clusterings must be somewhat similar to one another in order not to confuse the expert. We propose a CP model to build a partition that is similar to an existing partition while taking constraints into account. Our model can support several types of constraints and features constraint relaxation and cluster creation. Experiments performed on popular datasets as well as a use case in remote sensing denote the relevance of our model.
Document type :
Conference papers
Complete list of metadata

https://hal.archives-ouvertes.fr/hal-03687871
Contributor : Samir Loudni Connect in order to contact the contributor
Submitted on : Wednesday, September 28, 2022 - 2:30:31 PM
Last modification on : Thursday, December 1, 2022 - 10:54:03 AM

File

JFPC_2022.pdf
Files produced by the author(s)

Identifiers

  • HAL Id : hal-03687871, version 1

Citation

Aymeric Beauchamp, Thi-Bich-Hanh Dao, Samir Loudni, Christel Vrain. Intégration incrémentale de contraintes pour le clustering avec la programmation par contraintes. Journées Francophones de Programmation par Contraintes (Evènement affilié à PFIA 2022), Jun 2022, Saint-Étienne, France. ⟨hal-03687871⟩

Share

Metrics

Record views

55

Files downloads

0