In this workshop we propose to address the fundamental question of how the brain makes statistical inference and selects the most suitable models so as to act in the world. Since the pioneer work of Helmholtz (1867), it has been conjectured that the brain does statistical model selection by assigning probabilistic models to samples of stimuli and making statistical inference to predict upcoming events. This conjecture has been extensively discussed in the literature under several labels (predictive coding, Bayesian brain, etc). This multiplicity of labels blurs communication in the field. One of the goals of this workshop is to put together experts in the field towards a unifying conceptual framework.

Results supporting the conjecture that the brain effectively identifies the structure of the chain generating the sequence of stimuli were obtained by many labs. In particular, electrophysiological and behavioral signatures of impending and performed actions as well as signatures associated to sequence learning have been identified in the human brain. In another line of evidence, signatures of offline rehearsal through the process of learning motor sequences have been associated to performance improvement (i.e., the phenomenon of neural replay). Sequence learning also implies in implementing motivational drive models by the brain through reinforcement learning.

Discussing these fundamental issues will require the development of new experimental protocols, new mathematical models and new statistical methods. Besides these fundamental aspects, the workshop will discuss clinical applications and innovation in social sciences.