I have a dataset where I have around 50 independent variables used to run survival analysis on the target variables. But out of these 50 variables, 46 variables are categorical variables i.e. having values in range 0-5(Its a survey dataset where user has to mark the experience between 0 to 5 values, where 5 means the highest). Now, the survey has 10 questions which are to filled from time to time(5 times) in order to assess the change in the user experience. So I have the dataset containing the values for same 10 questions(different values) repeated 5 times to make it 50 columns.

My questions are:

  1. I want to do survival analysis on this dataset. Should I keep the dataset as it is(as the variables have different values from time to time) or should I process the dataset as the questions are repeated 5 times. If yes, can someone please advice any method to do so.
  2. I have tried cox regression and survival trees to learn a model. But the performance of these models is not so good. Also, I am performing feature selection but there is not much difference as almost all the feature have same characteristics.

Below is the image of few columns of the dataset. For privacy reasons, could not share the complete dataset. Kindly understand that almost all the variables of the dataset have such data type except for age and gender.

Any help would be appreciated. Thanks in advance!



1 Answer 1


It seems a problem that can be solved with a correct approach for survival analysis with time-dependent covariates.


This vignette is useful also for those not interested in R code.

Have you tried a random forest survival analysis approach?


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.