Description Details Author(s) References
Implementation in the density framework of the non-combinatorial algorithm and its greedy version, introduced by Magalhães and Rozenholc (2014), for T-estimation Hold-Out proposed in Birgé (2006, Section 9). The package provide an implementation which uses several families of estimators (regular and irregular histograms, kernel estimators) which may be used alone or combined. As a complement, provides also a comparison with other Held-Out derived from least-squares and maximum-likelihood.
Package: | Density.T.HoldOut |
Type: | Package |
Version: | 1.00 |
Date: | 2014-01-08 |
License: | GPL(>=2) |
Nelo Magalhães and Yves Rozenholc.
Maintainer: Nelo Magalhães <nelo.moltermagalhaes@gmail.com>
L. Birgé, "Model selection via testing: an alternative to (penalized) maximum likelihood estimators.", Ann. Institut Henri Poincaré Probab. et Statist., 42, 273–325, (2006)
N. Magalhães and Y. Rozenholc, "A non-combinatorial algorithm for T-estimation.", (2014)
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.