Journal article
THE Mnet METHOD FOR VARIABLE SELECTION
Statistica Sinica, Vol.26(3), pp.903-923
07/01/2016
DOI: 10.5705/ss.202014.0011
Abstract
We propose a penalized approach for variable selection using a combination of minimax concave and ridge penalties. The method is designed to deal with p >= n problems with highly correlated predictors. We call this the Mnet method. Similar to the elastic net of Zou and Hastie (2005), the Mnet tends to select or drop highly correlated predictors together. However, unlike the elastic net, the Mnet is selection consistent and equal to the oracle ridge estimator with high probability under reasonable conditions. We develop an efficient coordinate descent algorithm to compute the Mnet estimates. Simulation studies show that the Mnet has better performance in the presence of highly correlated predictors than either the elastic net or MCP. We illustrate the application of the Mnet to data from a gene expression study in ophthalmology.
Details
- Title: Subtitle
- THE Mnet METHOD FOR VARIABLE SELECTION
- Creators
- Jian Huang - University of IowaPatrick Breheny - University of IowaSangin Lee - The University of Texas Southwestern Medical CenterShuangge Ma - Yale UniversityCun-Hui Zhang - Rutgers, The State University of New Jersey
- Resource Type
- Journal article
- Publication Details
- Statistica Sinica, Vol.26(3), pp.903-923
- Publisher
- STATISTICA SINICA
- DOI
- 10.5705/ss.202014.0011
- ISSN
- 1017-0405
- eISSN
- 1996-8507
- Number of pages
- 21
- Grant note
- R01CA142774 / NIH DMS-1208225; DMS-1129626; DMS-1209014; DMS-1513378 / NSF H98230-15-1-0040 / NSA
- Language
- English
- Date published
- 07/01/2016
- Academic Unit
- Statistics and Actuarial Science; Biostatistics; Internal Medicine
- Record Identifier
- 9984257731702771
Metrics
11 Record Views