Séminaire du département Parole et Cognition du 18/10/2018 à 14h00

Neurophysiological and Computational Understanding of Entropy/Uncertainty encoding in Language- and Music-Statistical Learning: Towards Medical and Educational applications by a Brain-Computer Interfa

Intervenant : Tatsuya Daikoku, Max Planck Institute for Human Cognitive and Brain Sciences

Lieu : salle Chartreuse (D 1121)

Résumé : Statistical learning (SL), which is a learning system of transitional probabilities embedded in sequential phenomena such as music and language, has been considered an implicit and domain-general mechanism that is innate in human’s brain. The SL is also interdisciplinary notion including information technology, artificial intelligence, musicology, and linguistics as well as psychology and neuroscience. A line of recent studies suggest that information-theoretical notion of SL can be represented in neurophysiological responses in the framework of predictive coding. Here, I show a line of our neurophysiological and computational studies of SL in music and language. In addition, I discusses how statistically acquired knowledge is related to creativities of language, music, and motor activities. Then, I argues the promising approaches for the application of therapy and pedagogy, using a integrated method of neural and machine-learning approaches.