Variable selection and estimation using a continuous approximation to the \(L_0\) penalty (Q1695760)
From MaRDI portal
scientific article
| Language | Label | Description | Also known as |
|---|---|---|---|
| English | Variable selection and estimation using a continuous approximation to the \(L_0\) penalty |
scientific article |
Statements
Variable selection and estimation using a continuous approximation to the \(L_0\) penalty (English)
0 references
8 February 2018
0 references
In this paper, a new type of exponential penalty, the so-called EXP penalty, which closely resembles the \(L_0\) penalty, is proposed. The model selection oracle property is studied and the EXP is implemented using a coordinate descent (CD) algorithm. Also, a modified BIC (MBIC) tuning parameter selection method is proposed for EXP, and it is shown that it consistently identifies the correct model. Simulation results and a data example support the EXP procedure.
0 references
penalized least squares
0 references
coordinate descent algorithm
0 references
variable selection
0 references
MBIC
0 references
oracle property
0 references
0 references
0 references
0 references
