About this document
OSCAR for Generalized Linear Models by kaczetoww is a document available to read on EtoBox.
The document presents the OSCAR (Octagonal Selection and Clustering Algorithm in Regression) for Generalized Linear Models (GLMs), which allows for clustering of highly correlated predictors by assigning them the same coefficient. It discusses the OSCAR penalty region, estimation methods, and provides an algorithm to solve the corresponding maximization problem. The effectiveness of the OSCAR method is demonstrated through a simulation study and a real-world example from water engineering.
- Author
- kaczetoww
- Language
- EN