|
Abstract:
|
Latent process models are useful for modeling survival data but are even more useful for joint modeling several types of observations. Joint modeling appears to be a promising approach, for instance in the context of ageing, where it is possible to relate scores to psychometric tests and diagnosis of dementia to the same latent process representing cognitive ability. An observed event (for instance, dementia) may be determined by the latent process hitting a barrier; alternatively, the latent process may appear as a time-dependent variable in a proportional hazard model. We exhibit a general framework, relating observations (of events or of markers) to the latent process and making clear the difference between random transformations and coarsening.
|
