Section: New Results

Model based-clustering for pharmacovigilance data

Participants : Gilles Celeux, Christine Keribin, Valérie Robert.

In collaboration with Pascale Tubert-Bitter, Ismael Ahmed and Mohamed Sedki, Gilles Celeux and Christine Keribin have started research concerning the detection of associations between drugs and adverse events in the framework of the PhD of Valerie Robert. At first, this team developed a model-based clustering inspired by latent block models, which consists of co-clustering rows and columns of two binary tables, imposing the same row ranking. This enables it to highlight subgroups of individuals sharing the same drug profile, and subgroups of adverse effects and drugs with strong interactions. Furthermore, some sufficient conditions are provided to obtain the identifiability of the model, and some results are shown for simulated data. This year, the exact ICL criterion has been extended to this double block latent model. Moreover, the possible added value of this model, compared with standard contingency table analysis, is currently under scrutiny.