R help - Oct 2019 - L1 (lasso) regularized log-linear model selection procedure

Daphne Koller (2009) describes L1 regularization (Chapter 20) as an
efficient way for Markov network (i.e. undirected graphical model)
structure learning and feature parameter estimation.

Her focus, and mine, are log-linear models for high-dimensional
contingency tables (i.e. categorical data).

I wonder whether there are any good implementations of this?

I have looked here (https://cran.r-project.org/web/views/gR.html) and
found only implementations for continuous data:
* parcor: Regularized estimation of partial correlation matrices
* glasso: Graphical Lasso: Estimation of Gaussian Graphical Models

Both are for continuous (Gaussian) data, not categorical.

Any suggestions?

Searching on "lasso penalty with deviance" on rseek.org  brought up
many
packages.

-- Bert

Bert Gunter

"The trouble with having an open mind is that people keep coming along and
sticking things into it."
-- Opus (aka Berkeley Breathed in his "Bloom County" comic strip )


On Sat, Oct 19, 2019 at 7:54 AM Davor Josipovic <davorj at live.com>
wrote:
> Daphne Koller (2009) describes L1 regularization (Chapter 20) as an
> efficient way for Markov network (i.e. undirected graphical model)
> structure learning and feature parameter estimation.
>
> Her focus, and mine, are log-linear models for high-dimensional
> contingency tables (i.e. categorical data).
>
> I wonder whether there are any good implementations of this?
>
> I have looked here (https://cran.r-project.org/web/views/gR.html) and
> found only implementations for continuous data:
> * parcor: Regularized estimation of partial correlation matrices
> * glasso: Graphical Lasso: Estimation of Gaussian Graphical Models
>
> Both are for continuous (Gaussian) data, not categorical.
>
> Any suggestions?
> ______________________________________________
> R-help at r-project.org mailing list -- To UNSUBSCRIBE and more, see
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide
> http://www.R-project.org/posting-guide.html
> and provide commented, minimal, self-contained, reproducible code.
>
	[[alternative HTML version deleted]]