We will focus here on linear methods based on an orthogonal decomposition of the predictors

For the ridge estimator and the ordinary least squares estimator, and their variants, we provide new risk bounds of order d/n without logarithmic factor unlike some stan- dard

In order to satisfy criterion (4), in NIPALS, PLS-R and new Mode A PLS-PM, when work- ing on standardized variables, optimal weights are calculated as Pearson’s

We compute the decomposition numbers of the unipotent characters lying in the principal ℓ-block of a finite group of Lie type B 2n (q) or C 2n (q) when q is an odd prime power and ℓ

In this paper, we propose a novel approach to summarize the poste- rior distributions over variable-dimensional subspaces that typically arise in signal decomposition problems with

The copyright holder for this preprint which was not peer-reviewed is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.. It is made

In the proposed DLSP algorithm, the support of the solution is obtained by sorting the coefficients which are calcu- lated by the first Least-Squares, and then the non-zero values

Indeed, it generalizes the results of Birg´e and Massart (2007) on the existence of minimal penalties to heteroscedastic regression with a random design, even if we have to restrict

The results are expressed as the selected combination of criteria and methods with the number of components of a PLS regression which are close to the correct number of components