Nonparametric Stein-type shrinkage covariance matrix estimators in high-dimensional settings
- Submitting institution
-
University of Brighton
- Unit of assessment
- 11 - Computer Science and Informatics
- Output identifier
- 7123744
- Type
- D - Journal article
- DOI
-
10.1016/j.csda.2014.10.018
- Title of journal
- Computational Statistics & Data Analysis
- Article number
- -
- First page
- 251
- Volume
- 83
- Issue
- -
- ISSN
- 0167-9473
- Open access status
- Out of scope for open access requirements
- Month of publication
- October
- Year of publication
- 2014
- URL
-
-
- Supplementary information
-
-
- Request cross-referral to
- -
- Output has been delayed by COVID-19
- No
- COVID-19 affected output statement
- -
- Forensic science
- No
- Criminology
- No
- Interdisciplinary
- No
- Number of additional authors
-
0
- Research group(s)
-
-
- Citation count
- 24
- Proposed double-weighted
- No
- Reserve for an output with double weighting
- No
- Additional information
- Shrinkage approaches for estimating a high-dimensional covariance matrix are often employed to ensure that the estimated covariance matrix is non-singular and well-conditioned. This paper is significant because it proposes a new family of Stein-type shrinkage covariance matrix estimators that, unlike previous work: (a) can consider target matrices other than the scaled identity matrix, and (b) do not require a normality assumption about the underlying data generating process. A freely-available R package that calculates the estimators has been downloaded more than 24,000 times. The paper has inspired further theoretical developments including multiple shrinkage (Ikeda et al., Comput. Statist. Data Anal. 2016).
- Author contribution statement
- -
- Non-English
- No
- English abstract
- -