Pré-Publication, Document De Travail Année : 2023

Post-clustering Inference under Dependence

Résumé

Recent work by Gao et al. (JASA 2022) has laid the foundations for post-clustering inference, establishing a theoretical framework allowing to test for differences between means of estimated clusters. Additionally, they studied the estimation of unknown parameters while controlling the selective type I error. However, their theory was developed for independent observations identically distributed as p-dimensional Gaussian variables, where the parameter estimation could only be performed for spherical covariance matrices. Here, we aim at extending this framework to a more convenient scenario for practical applications, where arbitrary dependence structures between observations and features are allowed. We establish sufficient conditions for extending the setting presented by Gao et al. to the general dependence framework. Moreover, we assess theoretical conditions allowing the compatible estimation of a covariance matrix. The theory is developed for hierarchical agglomerative clustering algorithms with several types of linkages, and for the k-means algorithm. We illustrate our method with synthetic data and real data of protein structures.

Fichier principal
Vignette du fichier
PCIdep_2025.pdf (7.59 Mo) Télécharger le fichier
Origine Fichiers produits par l'(les) auteur(s)
Licence

Dates et versions

hal-04250364 , version 1 (19-10-2023)
hal-04250364 , version 2 (15-08-2025)

Licence

Identifiants

Citer

Javier González-Delgado, Mathis Deronzier, Juan Cortés, Pierre Neuvial. Post-clustering Inference under Dependence. 2025. ⟨hal-04250364v2⟩
509 Consultations
405 Téléchargements

Altmetric

Partager

  • More