Impact of Regularization in case of Logistic Regression

The goal of this workflow is to analyze the impact of different priors in case of the logistic regression. The workflow therefore first reads the internet advertisement dataset. Then it creates a subset with more columns than rows, favouring overfitting. In the next step three models with different prior options are trained. In the last step the results are summarized in an interactive javascript view.

External resources

• Regularization for Logistic Regression: L1, L2, Gauss or Laplace?