L*****2
发帖数: 66 | 1 I am building a survival model with few indepdent varaible being time
dependent. The values of these variables are not known in the future, for
instance, Bureau Score. In other words, when predicting the responses over
time starting at a certain point, we only know the values at that point,
not the values after that. My question is: should we use values at the
observation time for these variables or use the updated values over time in
the model development? |
|
