Skip to Main content Skip to Navigation
Journal articles

Estimating the number of block boundaries from diagonal blockwise matrices without penalization

Abstract : In computational biology, numerous recent studies have been dedicated to the analysis of the chromatin structure within the cell by two-dimensional segmentation methods. Motivated by this application, we consider the problem of retrieving the diagonal blocks in a matrix of observations. The theoretical properties of the least squares estimators of both the boundaries and the number of blocks are investigated. More precisely, the contribution of the paper is to establish the consistency of these estimators. A surprising consequence of our results is that, contrary to the one-dimensional case, a penalty is not needed for retrieving the true number of diagonal blocks. Finally, the results are illustrated on synthetic data.
Document type :
Journal articles
Complete list of metadatas

Cited literature [14 references]  Display  Hide  Download

https://hal.archives-ouvertes.fr/hal-01533843
Contributor : Archive Ouverte Prodinra <>
Submitted on : Tuesday, June 6, 2017 - 8:38:51 PM
Last modification on : Wednesday, September 16, 2020 - 5:26:21 PM
Long-term archiving on: : Thursday, September 7, 2017 - 2:18:35 PM

File

Brault_et_al-2017-Scandinavian...
Explicit agreement for this submission

Identifiers

Citation

Vincent Brault, Maud Delattre, Tristan Mary-Huard, Céline Leduc. Estimating the number of block boundaries from diagonal blockwise matrices without penalization. Scandinavian Journal of Statistics, Wiley, 2017, 44 (2), pp.563-580. ⟨10.1111/sjos.12266⟩. ⟨hal-01533843⟩

Share

Metrics

Record views

298

Files downloads

262