|
Abstract:
|
Logistic regression is a well known statistical tool aiming at linking some binary outcome to a set of explanatory variables. Recently, the idea has been extended to handle functional covariates, that is infinite-dimensional random elements, such as continuous-time stochastic processes for example. In this talk, we first give a brief overview of what the functional logistic model looks like and what are the main issues which arise when infinite-dimensional variables are involved. Then, we discuss new possible extensions of it, like incorporating some correlation between responses by including random effects in the model.
|
