MODELLING THE HUMAN BRAIN, WHILE IT MODELS THE SENSORY ENVIRONMENT

Garrido_all

Dr Marta Garrido, UQ 
This colloquium will be held at 1pm, 4th May, in Seddon 82D-301

Predictive Coding is a computational framework that posits the brain is a predictive, efficient, and adaptive machine. The ability to learn about regularities in the sensory environment and to make predictions about future events is fundamental for adaptive behaviour, as it may provide a competitive advantage for anticipating reward or avoiding punishment. In this talk, I will show how electroencephalography (EEG) and magnetoencephalography (MEG) combined with machine learning can be used to demonstrate that: 1) people are able to encode statistical regularities in the sensory environment even while their cognitive resources are heavily taxed; 2) violations to these regularities evoke sensory prediction errors that engage fronto-temporal networks with recurrent dynamic interactions; and finally, I will show 3) how the electrophysiology underlying predictive processes can be used to predict diagnosis of schizophrenia and psychotic traits in the general healthy population.