Dear Colleagues
I am trying to find an appropriate modelling technique whereby I can take
measurements from a particular individual over time, and then use this
behavioural data to predict whether that customer will reach a dichotomous
endpoint in the future. It seems like a logistic regression problem, but
the fact that I have repeated measurements over time will make this
approach invalid I feel???
Regards
Roger Westerbeek
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|