Bayesian Lasso with the Heredity Principle  

Oleh:

Noguchi, Hidehisa ; Ojima, Yoshikazu ; Yasui, Seiichi

Jenis: Article from Proceeding
Dalam koleksi: Asian Network for Quality (ANQ) Congress 2011, Ho Chi Minh City, Vietnam, 27-30 September 2011, page 1-9.
Topik: Variable Selection; Design of Experiments; Heredity Principle; HierarchicalModel; Penalized Regression
Fulltext: JP13_Hidehisa_Fullpaper.pdf (275.21KB)

Isi artikel

In the analysis of experiments, there are many variable selection methods for linear models. The lasso is a useful variable selection method and its estimates can be interpreted as a Bayesian posterior mode estimate when the regression parameters have independent Laplace priors. However the lasso method doesn’t consider the relationship between the predictors. In practice, this relationship is called as the effect heredity principle. In this paper, we extend the lasso with the heredity principle to Bayesian version. And we consider the relationship of posterior between the main effects and its corresponding interaction.

Opini Anda

Klik untuk menuliskan opini Anda tentang koleksi ini!

Kembali