# What is the correct way to report the results of a penalized (Ridge, LASSO, ElasticNet) logistic regression model?

I was wondering what would be a sensible way to report the results of a penalized logistic regression model in a scientific article (in terms of coefficients, metrics, diagnostics, hyperparameters, etc.)?
With an explanatory (unpenalized) model, I would report things like estimated coefficients, standard errors, degrees of freedom, (pseudo-)R2, p-values, test statistics, perhaps some 2D/3D plot... Here, I'm at a loss.

Some context: I'm working with an ElasticNet model (with glmnet and caret in R) but I guess answers apply equally to Ridge or LASSO models. My initial interest was to do inference. Yet, for various reasons (detailed here) I cannot use a "regular" logistic model to identify what explanatory variables drive my outcome so instead, I've decided to build a penalized logistic model in order to find values for my predictors beyond which the probabilities of my outcome cross some threshold values (and/or interpret coefficients directly) and make some quantified recommendations to obtain the desired outcome.

• What results do you mean? What do you want to present, and to whom will you present them? – Dave May 14 at 17:54
• What kind of information are you trying to learn by using the model? – Sycorax May 14 at 18:01
• You're both right, it was not very clear (I'm tired). I've edited my question to bring more context and precisions. I've also edited the linked question for it to be more focused in the hope someone will re-open it. Sorry for the inconvenience and thank you for your help. – Fanfoué May 14 at 18:56
• You say "My interest lies in inference." Yet when you say you want to "make some quantified recommendations to obtain the desired outcome," it sounds more like prediction. Inference is not easy with elastic net logistic regression; presentation of a predictive model can be much more straightforward. – EdM May 14 at 22:07
• @EdM, I initially wanted to do inference but, unfortunately, it was impossible. I changed my sentence a little bit to improve clarity on this point. I hear you when you say that presenting a predictive model can be much more straightforward, but how? What would you typically report? Only estimated coefficients with no measures of variability/uncertainty? – Fanfoué May 15 at 11:29

The best way to proceed--both from your and your readers' perspectives--is to repeat your entire modeling process on multiple bootstrap samples of your dataset (re-samples with replacement of the same size as your dataset) and report a summary of those results. That's explained in the context of LASSO for linear regression in Section 6.2 of Statistical Learning with Sparsity, but directly extends to elastic net for logistic regression if you use an appropriate measure for cross-validated selection of $$\alpha$$ (the ridge versus LASSO tradeoff) and $$\lambda$$ (penalization), like deviance (log loss). Don't rely on improper scoring rules like accuracy for building your model.