The study of separable data sets when Generalized Linear Model(GLM) is applied. Predictors that can completely predict the response of the model are called perfect predictors and lead to multiple problems in model inference, prediction, and also computation required for fitting the model. Approaches that deal with this problem fall into different categories, which are studied in this project. The goal of this study is to establish a recommendation system for which approach to be used based on fitting goals and exterior informations that we may have about the data.
SahandFarhoodi/Perfect_Predictors
Folders and files
| Name | Name | Last commit date | ||
|---|---|---|---|---|