The dataset is about events of death event as categorical variable(classifier) Since my variable of inteserest is survival, I have transformed the variable as survival (1 - Death) I have tried both univariate approach and stepwise regression method to fit the model In both approach, I have found **** as the significant variable Regression method: Binomial Logistic Regression (Since the outcome has only two catogory)
-
Notifications
You must be signed in to change notification settings - Fork 0
esksda/Survival-of-the-Burn-Injury-Patient
About
Survival probability of burn injury patients based on age, sex, race, hospital facility, and other significant facility nearing respondents.
Topics
Resources
Stars
Watchers
Forks
Releases
No releases published
Packages 0
No packages published