Accès ouvert · CC BY

A New Covariance Estimator for Sufficient Dimension Reduction in High-Dimensional and Undersized Sample Problems

Article scientifique 2022 Sciences mathématiques et statistiques Anglais

Résumé

Abstract The application of standard-sufficient dimension reduction (SDR) methods for reducing the dimension space of predictors without losing regression information requires inverting the covariance matrix of the predictors. This requirement has plagued their success in high-dimensional data analysis. We propose a unified regression-type estimation strategy, which replaces the degenerate maximum likelihood covariance matrix with a new covariance estimator called Maximum Entropy Covariance (MEC) estimator at the dimension reduction step to produce accurate solutions. The MEC deals directly with singularity and instability of sample group covariance estimates and addresses loss of covariance information in the high-dimensional SDR applications using Maximum Entropy (ME) principle for the first time. We demonstrate the effectiveness of the proposed MEC-SDR methods using real-life Leukemia cancer and customers' electricity load profiles from smart meter data sets.

Citer ce document

Olorede, K. O., & Yahya, W. B. (2022). A New Covariance Estimator for Sufficient Dimension Reduction in High-Dimensional and Undersized Sample Problems. Research Square. https://doi.org/10.21203/rs.3.rs-2206652/v1

Exporter : BibTeX · RIS (Zotero, Mendeley, EndNote)

Accès au document

Texte intégral archivé sur SPHAERO, en accès ouvert : lisible gratuitement en ligne après connexion. Le document reste aussi disponible à sa source. Se connecter

Voir l'article sur le site de la revue

Licence et provenance

Licence : CC BY

Notice moissonnée depuis OpenAlex le 09/09/2026. Le document reste hébergé par sa source.
Voir le document à la source →

Statistiques

Consultations (hors robots) : 10

Téléchargements (hors robots) : 0

Méthode de comptage