Can anyone explain how is Kaplan- Mier model a non parametric analogue of Binomial model and Nelson-Aalen model a non parametric analogue of Poisson Model?
If I remember correctly, K-M method does not allow the new increments during the experiment period, hence number of trials are fixed in advance : Binomial setting. N-A method does allow it, hence the number of experiments not fixed in advanced, and may increase. This suits the Poisson framework.
