MARS: selecting basis functions and knots with an empirical Bayes method (Q964645)
From MaRDI portal
scientific article
| Language | Label | Description | Also known as |
|---|---|---|---|
| English | MARS: selecting basis functions and knots with an empirical Bayes method |
scientific article |
Statements
MARS: selecting basis functions and knots with an empirical Bayes method (English)
0 references
22 April 2010
0 references
A Bayesian approach to multivariate adaptive regression spline (MARS) estimation is considered. It is shown that the Bayesian estimation is equivalent to maximum penalized likelihood estimation. Akaike's Bayes information criterion is used for the selection of the number and combination of basis functions. Applications to real data examples are also considered.
0 references
multivariate adaptive regression splines
0 references
penalized likelihood
0 references
Akaike information criterion
0 references
