A model for phyllochron analysis from interval censored measurements


Plant development consists in various synchronised processes.In particular, the process of leaf apperance orphyllochron, characterised by the time of appearance of each leaf, is often studied as an easily observed indicator. In this talk, I will present a statistical model for phyllochron with independent time intervals between leaves, when the times of leaf appearance are interval censored. . First, assuming Gaussian distribution of time intervals, we have implemented a Monte Carlo EM algorithm, which makes use of a generator from the truncated multivariate normal distribution. A more general parametrisation can be considered using the framework of semi-Markov models ; then the forward-backward algorithm, originally developed for hidden semi-Markov models, can be adapted to the interval censoring context.​